{
  "title": "2026-09-25 AI Daily | Gemini 3.8 Live Released, Real-time Multimodal and Agent Observability Emerge as New Focus Points",
  "url": "https://miaok.ong/en/ai-daily/ai-daily-2026-09-25/",
  "date": "2026-09-25T07:00:00+08:00",
  "lastmod": "2026-09-25T07:00:00+08:00",
  "type": "ai-daily",
  "kind": "page",
  "language": "en",
  "description": "Google DeepMind launched Gemini 3.8 Live and Live Avatar, pushing real-time multimodal interaction to the infrastructure layer; agents are moving from planning to collaboration and observability, with Bot Screen and Claude Code cloud conversation response transparency and scalability needs; synthetic characters predicting negative outcomes for real audiences and the Bitget theft incident serve as a reminder that validity and risk control verification are more critical than model capabilities.",
  "keywords": null,
  "tags": [],
  "categories": [],
  "author": "Mark (Miao) Kong",
  "image": "https://miaok.ong/images/avatar.jpg",
  "content": "\u003ch1 id=\"2026-09-25-ai-daily--gemini-38-live-launches-with-a-new-focus-on-real-time-multimodality-and-agent-observability\"\u003e\n  2026-09-25 AI Daily | Gemini 3.8 Live Launches, with a New Focus on Real-time Multimodality and Agent Observability\n  \u003ca class=\"heading-link\" href=\"#2026-09-25-ai-daily--gemini-38-live-launches-with-a-new-focus-on-real-time-multimodality-and-agent-observability\"\u003e\n    \u003ci class=\"fa-solid fa-link\" aria-hidden=\"true\" title=\"Link to heading\"\u003e\u003c/i\u003e\n    \u003cspan class=\"sr-only\"\u003eLink to heading\u003c/span\u003e\n  \u003c/a\u003e\n\u003c/h1\u003e\n\u003cblockquote\u003e\n\u003cp\u003eGoogle DeepMind launches Gemini 3.8 Live and Live Avatar, pushing real-time multimodal interaction to the infrastructure level. Agents are evolving from planning to collaboration and observability, with tools like Bot Screen and Claude Code\u0026rsquo;s cloud sessions addressing the need for transparency and scalability. Meanwhile, the negative results of using synthetic personas to predict real audience outcomes and the Bitget hacking incident serve as reminders that effectiveness and risk control validation are more critical than model capabilities alone.\u003c/p\u003e\n\u003c/blockquote\u003e\n\u003ch2 id=\"-this-issues-watch-list-a-deep-dive\"\u003e\n  📖 This Issue\u0026rsquo;s Watch List: A Deep Dive\n  \u003ca class=\"heading-link\" href=\"#-this-issues-watch-list-a-deep-dive\"\u003e\n    \u003ci class=\"fa-solid fa-link\" aria-hidden=\"true\" title=\"Link to heading\"\u003e\u003c/i\u003e\n    \u003cspan class=\"sr-only\"\u003eLink to heading\u003c/span\u003e\n  \u003c/a\u003e\n\u003c/h2\u003e\n\u003cp\u003eThere are three main themes worth watching today. First, multimodality and real-time interaction: Google DeepMind\u0026rsquo;s launch of Gemini 3.8 Live and Live Avatar brings real-time conversation and interactive avatars into the infrastructure discussion; Ovis-Embedding uses a unified backbone to integrate text, images, video, and audio, and discussions around Claude Opus 5.5\u0026rsquo;s capabilities are also worth noting. Second, the shift of agents from planning to collaboration: X-Planner adds event structure planning for embodied intelligence, ReAdapt explores social relationship reasoning in LLMs, and COMED discusses the middle ground of multi-model routing and collaboration, making it a deep read for teams focused on agent architecture. Third, validation issues in vertical-specific scenarios: Research such as MedGate-Fusion, explainable GBM for survival prediction, and the negative results of using synthetic personas to predict real audience outcomes highlights that \u0026ldquo;effectiveness validation\u0026rdquo; is more crucial than pure model capability when implementing solutions in healthcare and marketing.\u003c/p\u003e\n\u003ch2 id=\"-ai-hot-topics-on-x\"\u003e\n  🌐 AI Hot Topics on X\n  \u003ca class=\"heading-link\" href=\"#-ai-hot-topics-on-x\"\u003e\n    \u003ci class=\"fa-solid fa-link\" aria-hidden=\"true\" title=\"Link to heading\"\u003e\u003c/i\u003e\n    \u003cspan class=\"sr-only\"\u003eLink to heading\u003c/span\u003e\n  \u003c/a\u003e\n\u003c/h2\u003e\n\u003ch3 id=\"topic-1-nous-research-launches-bot-screen-for-live-ai-agent-monitoring\"\u003e\n  Topic 1: Nous Research Launches Bot Screen for Live AI Agent Monitoring\n  \u003ca class=\"heading-link\" href=\"#topic-1-nous-research-launches-bot-screen-for-live-ai-agent-monitoring\"\u003e\n    \u003ci class=\"fa-solid fa-link\" aria-hidden=\"true\" title=\"Link to heading\"\u003e\u003c/i\u003e\n    \u003cspan class=\"sr-only\"\u003eLink to heading\u003c/span\u003e\n  \u003c/a\u003e\n\u003c/h3\u003e\n\u003cul\u003e\n\u003cli\u003eCategory: AI · News\u003c/li\u003e\n\u003cli\u003eOverview: Trending for: 1 day ago, Related posts: 842\u003c/li\u003e\n\u003cli\u003eWhat it is: Nous Research has launched Bot Screen, a tool for real-time monitoring of AI agent operational status.\u003c/li\u003e\n\u003cli\u003eWhy it\u0026rsquo;s important: As the number of autonomous AI agents grows, real-time observability has become critical for secure deployment and debugging. This tool addresses the industry\u0026rsquo;s demand for agent transparency.\u003c/li\u003e\n\u003cli\u003eDiscussion summary: Discussions on X are focused on whether the tool can effectively reduce the risk of AI agents going rogue, how it differs from existing monitoring solutions, and the impact of real-time monitoring on agent privacy and performance.\u003c/li\u003e\n\u003c/ul\u003e\n\u003ch3 id=\"topic-2-bitget-crypto-exchange-hit-by-3516-million-hot-wallet-breach\"\u003e\n  Topic 2: Bitget Crypto Exchange Hit by $351.6 Million Hot Wallet Breach\n  \u003ca class=\"heading-link\" href=\"#topic-2-bitget-crypto-exchange-hit-by-3516-million-hot-wallet-breach\"\u003e\n    \u003ci class=\"fa-solid fa-link\" aria-hidden=\"true\" title=\"Link to heading\"\u003e\u003c/i\u003e\n    \u003cspan class=\"sr-only\"\u003eLink to heading\u003c/span\u003e\n  \u003c/a\u003e\n\u003c/h3\u003e\n\u003cul\u003e\n\u003cli\u003eCategory: AI · News\u003c/li\u003e\n\u003cli\u003eOverview: Trending for: 3 hours ago, Related posts: 16,000\u003c/li\u003e\n\u003cli\u003eWhat it is: The crypto exchange Bitget suffered an attack on its hot wallet, resulting in a loss of approximately $351.6 million.\u003c/li\u003e\n\u003cli\u003eWhy it\u0026rsquo;s important: This incident highlights the vulnerabilities of digital asset platforms in real-time risk control and anomaly detection. It may accelerate the adoption of AI-driven on-chain monitoring, automated threat response, and security audit technologies.\u003c/li\u003e\n\u003cli\u003eDiscussion summary: The main topics of discussion on X are the attack methods and the platform\u0026rsquo;s responsibility. The focus is on whether user funds were affected, the platform\u0026rsquo;s ability to pay, and whether AI can proactively identify abnormal transactions or counter increasingly sophisticated attacks.\u003c/li\u003e\n\u003c/ul\u003e\n\u003ch3 id=\"topic-3-house-report-accuses-doj-of-secretly-targeting-republican-lawmakers\"\u003e\n  Topic 3: House Report Accuses DOJ of Secretly Targeting Republican Lawmakers\n  \u003ca class=\"heading-link\" href=\"#topic-3-house-report-accuses-doj-of-secretly-targeting-republican-lawmakers\"\u003e\n    \u003ci class=\"fa-solid fa-link\" aria-hidden=\"true\" title=\"Link to heading\"\u003e\u003c/i\u003e\n    \u003cspan class=\"sr-only\"\u003eLink to heading\u003c/span\u003e\n  \u003c/a\u003e\n\u003c/h3\u003e\n\u003cul\u003e\n\u003cli\u003eCategory: AI · Other\u003c/li\u003e\n\u003cli\u003eOverview: Trending for: , Related posts: 125\u003c/li\u003e\n\u003cli\u003eWhat it is: Republicans on the U.S. House Judiciary Committee released declassified files from the FBI\u0026rsquo;s \u0026ldquo;Arctic Frost\u0026rdquo; investigation, accusing the Department of Justice (DOJ) of secretly monitoring Republican lawmakers and allies of Donald Trump.\u003c/li\u003e\n\u003cli\u003eWhy it\u0026rsquo;s important: The incident exposes large-scale data collection (financial, social media, and communications) by law enforcement agencies. This could intensify debates around AI-assisted surveillance, data privacy, and algorithmic accountability, potentially affecting the regulatory boundaries for government use of AI tools.\u003c/li\u003e\n\u003cli\u003eDiscussion summary: The focus on X is whether the DOJ has been weaponized. Supporters claim the evidence shows the FBI overstepped its authority in collaboration with the J6 Committee to target conservatives. Opponents argue it is a standard accountability measure related to interference in the 2020 election and the Capitol riot.\u003c/li\u003e\n\u003c/ul\u003e\n\u003ch3 id=\"topic-4-alexandr-wangs-camo-crocs-outfit-sparks-luxury-memes-at-meta-connect\"\u003e\n  Topic 4: Alexandr Wang\u0026rsquo;s Camo Crocs Outfit Sparks Luxury Memes at Meta Connect\n  \u003ca class=\"heading-link\" href=\"#topic-4-alexandr-wangs-camo-crocs-outfit-sparks-luxury-memes-at-meta-connect\"\u003e\n    \u003ci class=\"fa-solid fa-link\" aria-hidden=\"true\" title=\"Link to heading\"\u003e\u003c/i\u003e\n    \u003cspan class=\"sr-only\"\u003eLink to heading\u003c/span\u003e\n  \u003c/a\u003e\n\u003c/h3\u003e\n\u003cul\u003e\n\u003cli\u003eCategory: AI · News\u003c/li\u003e\n\u003cli\u003eOverview: Trending for: 17 hours ago, Related posts: 4,700\u003c/li\u003e\n\u003cli\u003eWhat it is: Alexandr Wang\u0026rsquo;s outfit at Meta Connect, featuring camouflage Crocs paired with formal wear, sparked a wave of luxury-brand-style memes created by netizens.\u003c/li\u003e\n\u003cli\u003eWhy it\u0026rsquo;s important: This event reflects how the personal image and public profile of star founders in the AI field can transcend technical discussions, becoming part of the industry\u0026rsquo;s cultural dissemination and brand identity.\u003c/li\u003e\n\u003cli\u003eDiscussion summary: Discussions centered on joking about the clash between his mix-and-match style and Silicon Valley\u0026rsquo;s elite dress code. People also debated whether these lighthearted memes help bridge the gap between AI leaders and the public or distract from important technical topics.\u003c/li\u003e\n\u003c/ul\u003e\n\u003ch3 id=\"topic-5-clm-model-speeds-up-ai-agent-decisions-with-contrastive-learning\"\u003e\n  Topic 5: CLM Model Speeds Up AI Agent Decisions with Contrastive Learning\n  \u003ca class=\"heading-link\" href=\"#topic-5-clm-model-speeds-up-ai-agent-decisions-with-contrastive-learning\"\u003e\n    \u003ci class=\"fa-solid fa-link\" aria-hidden=\"true\" title=\"Link to heading\"\u003e\u003c/i\u003e\n    \u003cspan class=\"sr-only\"\u003eLink to heading\u003c/span\u003e\n  \u003c/a\u003e\n\u003c/h3\u003e\n\u003cul\u003e\n\u003cli\u003eCategory: AI · News\u003c/li\u003e\n\u003cli\u003eOverview: Trending for: 20 hours ago, Related posts: 1400\u003c/li\u003e\n\u003cli\u003eWhat it is: A model named CLM has significantly increased the decision-making speed of AI agents through contrastive learning.\u003c/li\u003e\n\u003cli\u003eWhy it matters: This advancement is expected to reduce the inference latency of AI agents, improving efficiency in real-time interactions and automated tasks, and has positive implications for deployment in resource-constrained environments.\u003c/li\u003e\n\u003cli\u003eDiscussion summary: Discussions on X are mainly focused on the practical advantages of this model compared to traditional reinforcement learning methods, its generalization ability and stability in complex environments, and whether contrastive learning is genuinely effective for decision-making tasks.\u003c/li\u003e\n\u003c/ul\u003e\n\u003ch3 id=\"topic-6-anthropic-launches-cloud-sessions-for-claude-code-with-subscriber-credits\"\u003e\n  Topic 6: Anthropic Launches Cloud Sessions for Claude Code with Subscriber Credits\n  \u003ca class=\"heading-link\" href=\"#topic-6-anthropic-launches-cloud-sessions-for-claude-code-with-subscriber-credits\"\u003e\n    \u003ci class=\"fa-solid fa-link\" aria-hidden=\"true\" title=\"Link to heading\"\u003e\u003c/i\u003e\n    \u003cspan class=\"sr-only\"\u003eLink to heading\u003c/span\u003e\n  \u003c/a\u003e\n\u003c/h3\u003e\n\u003cul\u003e\n\u003cli\u003eCategory: AI · News\u003c/li\u003e\n\u003cli\u003eOverview: Trending for: 1 day ago, Related posts: 8000\u003c/li\u003e\n\u003cli\u003eWhat it is: Anthropic has launched a cloud sessions feature for Claude Code, allowing subscribers to use their account credits to run coding agents in the cloud.\u003c/li\u003e\n\u003cli\u003eWhy it matters: This marks the expansion of AI coding assistants from local command-line tools to managed cloud services, which could lower the barrier to entry for developers and promote the adoption of agent-based programming.\u003c/li\u003e\n\u003cli\u003eDiscussion summary: Discussions on X are primarily focused on the credit consumption rate and cost, the difference in experience between cloud sessions and local execution, security and latency concerns, and its positioning relative to competitors like OpenAI Codex and Cursor.\u003c/li\u003e\n\u003c/ul\u003e\n\u003ch3 id=\"topic-7-ai-agent-trims-400000-lines-of-redundant-tests-in-openclaw-project\"\u003e\n  Topic 7: AI Agent Trims 400,000 Lines of Redundant Tests in OpenClaw Project\n  \u003ca class=\"heading-link\" href=\"#topic-7-ai-agent-trims-400000-lines-of-redundant-tests-in-openclaw-project\"\u003e\n    \u003ci class=\"fa-solid fa-link\" aria-hidden=\"true\" title=\"Link to heading\"\u003e\u003c/i\u003e\n    \u003cspan class=\"sr-only\"\u003eLink to heading\u003c/span\u003e\n  \u003c/a\u003e\n\u003c/h3\u003e\n\u003cul\u003e\n\u003cli\u003eCategory: AI · News\u003c/li\u003e\n\u003cli\u003eOverview: Trending for: 6 hours ago, Related posts: 367\u003c/li\u003e\n\u003cli\u003eWhat it is: An AI agent cleaned up approximately 400,000 lines of redundant test code in the OpenClaw project.\u003c/li\u003e\n\u003cli\u003eWhy it matters: This demonstrates the capability of AI agents in large-scale code cleanup and automated maintenance, potentially reducing technical debt and improving development efficiency significantly.\u003c/li\u003e\n\u003cli\u003eDiscussion summary: The discussion mainly revolves around the accuracy of the AI cleanup, whether it might mistakenly delete valid tests, and if such automated operations should undergo more rigorous manual review.\u003c/li\u003e\n\u003c/ul\u003e\n\u003ch3 id=\"topic-8-paper-rex-beats-team-liquid-2-1-to-top-valorant-champions-odds\"\u003e\n  Topic 8: Paper Rex Beats Team Liquid 2-1 to Top VALORANT Champions Odds\n  \u003ca class=\"heading-link\" href=\"#topic-8-paper-rex-beats-team-liquid-2-1-to-top-valorant-champions-odds\"\u003e\n    \u003ci class=\"fa-solid fa-link\" aria-hidden=\"true\" title=\"Link to heading\"\u003e\u003c/i\u003e\n    \u003cspan class=\"sr-only\"\u003eLink to heading\u003c/span\u003e\n  \u003c/a\u003e\n\u003c/h3\u003e\n\u003cul\u003e\n\u003cli\u003eCategory: AI · Sports\u003c/li\u003e\n\u003cli\u003eOverview: Trending for:, Related posts: 164\u003c/li\u003e\n\u003cli\u003eWhat it is: Paper Rex defeated Team Liquid 2-1, rising to become the top favorite in the VALORANT Champions betting odds.\u003c/li\u003e\n\u003cli\u003eWhy it matters: This event has garnered attention as a topic at the intersection of AI and sports, reflecting how the influence and accuracy of AI-driven match predictions and betting models in esports are being put to a real-world test.\u003c/li\u003e\n\u003cli\u003eDiscussion summary: Discussions on X mainly center on whether AI predictions are overestimating Paper Rex, the reasonableness of the odds changes, and the credibility of data models in esports forecasting.\u003c/li\u003e\n\u003c/ul\u003e\n\u003ch3 id=\"topic-9-israeli-forward-scores-then-sent-off-for-flag-celebration\"\u003e\n  Topic 9: Israeli Forward Scores Then Sent Off for Flag Celebration\n  \u003ca class=\"heading-link\" href=\"#topic-9-israeli-forward-scores-then-sent-off-for-flag-celebration\"\u003e\n    \u003ci class=\"fa-solid fa-link\" aria-hidden=\"true\" title=\"Link to heading\"\u003e\u003c/i\u003e\n    \u003cspan class=\"sr-only\"\u003eLink to heading\u003c/span\u003e\n  \u003c/a\u003e\n\u003c/h3\u003e\n\u003cul\u003e\n\u003cli\u003eCategory: AI · Other\u003c/li\u003e\n\u003cli\u003eOverview: Trending for: 3 hours ago, Related posts: 17000\u003c/li\u003e\n\u003cli\u003eWhat it is: An Israeli football forward was sent off with a red card by the referee for celebrating a goal by displaying a flag.\u003c/li\u003e\n\u003cli\u003eWhy it matters: While the event itself is not directly related to AI technology, its automatic categorization under \u0026ldquo;AI · Other\u0026rdquo; and its high trending status reflect the challenges of automated content classification, politically sensitive content identification, and recommendation algorithms in scenarios where sports and politics intersect. This serves as a case study for AI content moderation.\u003c/li\u003e\n\u003cli\u003eDiscussion summary: Discussions on X are mainly focused on whether the red card was justified, whether the flag constitutes a political statement, and the boundaries of political symbolism in sporting events. Users are divided on refereeing standards and the platform\u0026rsquo;s handling of content labeling.\u003c/li\u003e\n\u003c/ul\u003e\n\u003ch3 id=\"topic-10-dagestani-fighters-relentless-discipline-powers-ufc-dominance\"\u003e\n  Topic 10: Dagestani Fighters\u0026rsquo; Relentless Discipline Powers UFC Dominance\n  \u003ca class=\"heading-link\" href=\"#topic-10-dagestani-fighters-relentless-discipline-powers-ufc-dominance\"\u003e\n    \u003ci class=\"fa-solid fa-link\" aria-hidden=\"true\" title=\"Link to heading\"\u003e\u003c/i\u003e\n    \u003cspan class=\"sr-only\"\u003eLink to heading\u003c/span\u003e\n  \u003c/a\u003e\n\u003c/h3\u003e\n\u003cul\u003e\n\u003cli\u003eCategory: AI · Other\u003c/li\u003e\n\u003cli\u003eOverview: Trending for: 8 hours ago, Related posts: 328\u003c/li\u003e\n\u003cli\u003eWhat it is: Dagestani fighters have achieved a dominant position in the UFC through their rigorous discipline.\u003c/li\u003e\n\u003cli\u003eWhy it matters: Although this topic is from the realm of mixed martial arts, it has drawn attention from the AI community to the methodology of \u0026ldquo;discipline-driven iteration\u0026rdquo; due to the analogy between their training discipline and the principle of continuous optimization in reinforcement learning.\u003c/li\u003e\n\u003cli\u003eDiscussion summary: Discussions on X are centered on whether the training system of Dagestani fighters can be analogized to the design of reward functions in AI, and whether extreme discipline from non-technical fields can offer insights for the long-term, robust training of AI models.\u003c/li\u003e\n\u003c/ul\u003e\n\u003ch3 id=\"topic-11-van-dijk-commits-to-netherlands-as-gakpos-late-goal-earns-draw-vs-klopps-germany\"\u003e\n  Topic 11: Van Dijk Commits to Netherlands as Gakpo\u0026rsquo;s Late Goal Earns Draw vs. Klopp\u0026rsquo;s Germany\n  \u003ca class=\"heading-link\" href=\"#topic-11-van-dijk-commits-to-netherlands-as-gakpos-late-goal-earns-draw-vs-klopps-germany\"\u003e\n    \u003ci class=\"fa-solid fa-link\" aria-hidden=\"true\" title=\"Link to heading\"\u003e\u003c/i\u003e\n    \u003cspan class=\"sr-only\"\u003eLink to heading\u003c/span\u003e\n  \u003c/a\u003e\n\u003c/h3\u003e\n\u003cul\u003e\n\u003cli\u003eCategory: AI · Other\u003c/li\u003e\n\u003cli\u003eOverview: Trending for: 1 day ago, Related posts: 57000\u003c/li\u003e\n\u003cli\u003eWhat it is: The Netherlands drew with Klopp\u0026rsquo;s Germany thanks to a last-minute goal from Gakpo, and Van Dijk has confirmed he will continue to play for the Dutch national team.\u003c/li\u003e\n\u003cli\u003eWhy it matters: This topic is a sports hotspot with no direct technical or business impact on the AI field. Its inclusion in the \u0026ldquo;AI · Other\u0026rdquo; trend reflects the vague boundaries of social media topic categorization for non-AI content, which may interfere with the observation of AI trends.\u003c/li\u003e\n\u003cli\u003eDiscussion summary: Discussions on X primarily revolved around Van Dijk\u0026rsquo;s future with the national team, Gakpo\u0026rsquo;s crucial goal, and Klopp\u0026rsquo;s tactical adjustments after becoming Germany\u0026rsquo;s coach. There was no significant discussion or disagreement on this topic within the AI community.\u003c/li\u003e\n\u003c/ul\u003e\n\u003ch3 id=\"topic-12-lewis-hamilton-shares-ai-video-of-f1-drivers-in-funny-car-ride\"\u003e\n  Topic 12: Lewis Hamilton Shares AI Video of F1 Drivers in Funny Car Ride\n  \u003ca class=\"heading-link\" href=\"#topic-12-lewis-hamilton-shares-ai-video-of-f1-drivers-in-funny-car-ride\"\u003e\n    \u003ci class=\"fa-solid fa-link\" aria-hidden=\"true\" title=\"Link to heading\"\u003e\u003c/i\u003e\n    \u003cspan class=\"sr-only\"\u003eLink to heading\u003c/span\u003e\n  \u003c/a\u003e\n\u003c/h3\u003e\n\u003cul\u003e\n\u003cli\u003eCategory: AI · Sports\u003c/li\u003e\n\u003cli\u003eOverview: Trending for: 8 hours ago, Related posts: 10000\u003c/li\u003e\n\u003cli\u003eWhat happened: Lewis Hamilton shared a humorous AI-generated video on X featuring F1 drivers in a car.\u003c/li\u003e\n\u003cli\u003eWhy it matters: This event demonstrates the viral potential of generative AI in sports entertainment content and raises concerns about the authenticity and boundaries of AI-generated videos.\u003c/li\u003e\n\u003cli\u003eDiscussion summary: The discussion focused on the video\u0026rsquo;s realism and entertainment value, as well as issues such as whether AI-synthesized content could mislead viewers and potential violations of drivers\u0026rsquo; image rights.\u003c/li\u003e\n\u003c/ul\u003e\n\u003ch3 id=\"topic-13-chiles-golden-retriever-puppies-nap-after-military-parade\"\u003e\n  Topic 13: Chile\u0026rsquo;s Golden Retriever Puppies Nap After Military Parade\n  \u003ca class=\"heading-link\" href=\"#topic-13-chiles-golden-retriever-puppies-nap-after-military-parade\"\u003e\n    \u003ci class=\"fa-solid fa-link\" aria-hidden=\"true\" title=\"Link to heading\"\u003e\u003c/i\u003e\n    \u003cspan class=\"sr-only\"\u003eLink to heading\u003c/span\u003e\n  \u003c/a\u003e\n\u003c/h3\u003e\n\u003cul\u003e\n\u003cli\u003eCategory: AI · Entertainment\u003c/li\u003e\n\u003cli\u003eOverview: Trending for: , Related posts: 146\u003c/li\u003e\n\u003cli\u003eWhat happened: After a military parade in Chile, videos and photos of several Golden Retriever puppies napping on the roadside went viral on X and were categorized by the platform under the \u0026ldquo;AI · Entertainment\u0026rdquo; trend.\u003c/li\u003e\n\u003cli\u003eWhy it matters: This topic reflects how AI recommendation systems automatically cluster and amplify entertaining pet content. It also sparked discussions about the difficulty of distinguishing between real animal videos and AI-generated content.\u003c/li\u003e\n\u003cli\u003eDiscussion summary: Users on X primarily debated whether the puppy footage was real or AI-generated, and why the platform\u0026rsquo;s algorithm placed natural pet content into an AI category. Others focused on the impact of AI recommendation mechanisms on the spread of entertainment topics.\u003c/li\u003e\n\u003c/ul\u003e\n\u003ch3 id=\"topic-14-birdman-and-toni-braxton-spotted-together-after-rocky-marriage-timeline\"\u003e\n  Topic 14: Birdman and Toni Braxton Spotted Together After Rocky Marriage Timeline\n  \u003ca class=\"heading-link\" href=\"#topic-14-birdman-and-toni-braxton-spotted-together-after-rocky-marriage-timeline\"\u003e\n    \u003ci class=\"fa-solid fa-link\" aria-hidden=\"true\" title=\"Link to heading\"\u003e\u003c/i\u003e\n    \u003cspan class=\"sr-only\"\u003eLink to heading\u003c/span\u003e\n  \u003c/a\u003e\n\u003c/h3\u003e\n\u003cul\u003e\n\u003cli\u003eCategory: AI · Entertainment\u003c/li\u003e\n\u003cli\u003eOverview: Trending for: , Related posts: 48\u003c/li\u003e\n\u003cli\u003eWhat happened: Birdman and Toni Braxton were spotted together again after multiple breakups and reconciliations, a period the media has dubbed a \u0026ldquo;rocky marriage timeline.\u0026rdquo;\u003c/li\u003e\n\u003cli\u003eWhy it matters: This topic\u0026rsquo;s entry into the AI · Entertainment trend list shows that social media AI recommendation and auto-classification systems are including high-engagement entertainment content under AI-related tags. This exposes issues with cross-domain content understanding and tagging accuracy, serving as a classic case for evaluating bias in AI content moderation and classification.\u003c/li\u003e\n\u003cli\u003eDiscussion summary: Users on X mainly debated whether the two had gotten back together or secretly married, with fans reviewing their relationship timeline. Some also questioned the rationality of classifying this entertainment gossip as AI-related, discussing why AI tags failed to distinguish between entertainment and technology topics.\u003c/li\u003e\n\u003c/ul\u003e\n\u003ch3 id=\"topic-15-onlyfans-faces-claims-of-fake-subscribers-and-hidden-motives\"\u003e\n  Topic 15: OnlyFans Faces Claims of Fake Subscribers and Hidden Motives\n  \u003ca class=\"heading-link\" href=\"#topic-15-onlyfans-faces-claims-of-fake-subscribers-and-hidden-motives\"\u003e\n    \u003ci class=\"fa-solid fa-link\" aria-hidden=\"true\" title=\"Link to heading\"\u003e\u003c/i\u003e\n    \u003cspan class=\"sr-only\"\u003eLink to heading\u003c/span\u003e\n  \u003c/a\u003e\n\u003c/h3\u003e\n\u003cul\u003e\n\u003cli\u003eCategory: AI · Entertainment\u003c/li\u003e\n\u003cli\u003eOverview: Trending for: , Related posts: 50\u003c/li\u003e\n\u003cli\u003eWhat happened: OnlyFans has been accused of having a large number of fake subscribers and allegedly hiding the platform\u0026rsquo;s operational motives.\u003c/li\u003e\n\u003cli\u003eWhy it matters: The incident highlights how generative AI and automated bots can be used to forge users and interactions, exacerbating the platform\u0026rsquo;s trust crisis and impacting the creator economy.\u003c/li\u003e\n\u003cli\u003eDiscussion summary: Discussions on X focused on whether the fake accounts were AI-driven, whether the platform condoned or participated in this activity, and the potential impact on creators\u0026rsquo; income and the content ecosystem.\u003c/li\u003e\n\u003c/ul\u003e\n\u003ch4 id=\"summary-of-ai-public-opinion-on-x-today\"\u003e\n  Summary of AI Public Opinion on X Today\n  \u003ca class=\"heading-link\" href=\"#summary-of-ai-public-opinion-on-x-today\"\u003e\n    \u003ci class=\"fa-solid fa-link\" aria-hidden=\"true\" title=\"Link to heading\"\u003e\u003c/i\u003e\n    \u003cspan class=\"sr-only\"\u003eLink to heading\u003c/span\u003e\n  \u003c/a\u003e\n\u003c/h4\u003e\n\u003cp\u003eToday\u0026rsquo;s main discussion revolved around \u0026ldquo;trust and boundaries\u0026rdquo; as AI agents and generative AI accelerate their implementation. The industry generally agrees that real-time observability, low-latency decision-making, cloud-based agent coding, and automated code cleanup are key to enhancing the security and efficiency of AI deployment. However, the Bitget hack and the OnlyFans fake subscriber incident reveal that AI can be used not only for risk control but also for attacks or faking interactions, leading to significant disagreement on whether risks can be effectively prevented. In terms of content governance, multiple sports and entertainment topics being algorithmically classified as AI, combined with the authenticity debates over the F1 parody video and cute pet videos, reflect the unreliability of recommendation and moderation systems in cross-domain understanding and identifying AI-generated content. Furthermore, the FBI surveillance files have intensified the political opposition regarding government use of AI-assisted monitoring and privacy boundaries. Potential risks are concentrated in autonomous agents losing control or making erroneous deletions, automated fraud in finance and the creator economy, misinformation from AI-synthesized content, and the potential for inaccurate platform algorithm classification to amplify governance chaos.\u003c/p\u003e\n\u003ch2 id=\"-influencer-insights\"\u003e\n  💡 Influencer Insights\n  \u003ca class=\"heading-link\" href=\"#-influencer-insights\"\u003e\n    \u003ci class=\"fa-solid fa-link\" aria-hidden=\"true\" title=\"Link to heading\"\u003e\u003c/i\u003e\n    \u003cspan class=\"sr-only\"\u003eLink to heading\u003c/span\u003e\n  \u003c/a\u003e\n\u003c/h2\u003e\n\u003cblockquote\u003e\n\u003cp\u003eNo influencer insights for today. We recommend reading the in-depth content from the Watch List.\u003c/p\u003e\n\u003c/blockquote\u003e\n\u003ch2 id=\"-appendix-todays-watch-list-update-source-list\"\u003e\n  📚 Appendix: Today\u0026rsquo;s Watch List Update Source List\n  \u003ca class=\"heading-link\" href=\"#-appendix-todays-watch-list-update-source-list\"\u003e\n    \u003ci class=\"fa-solid fa-link\" aria-hidden=\"true\" title=\"Link to heading\"\u003e\u003c/i\u003e\n    \u003cspan class=\"sr-only\"\u003eLink to heading\u003c/span\u003e\n  \u003c/a\u003e\n\u003c/h2\u003e\n\u003cblockquote\u003e\n\u003cp\u003eTime window: Last 3 days; covers 22 sources; 33 updates in total\u003c/p\u003e\n\u003c/blockquote\u003e\n\u003ch3 id=\"stratechery-by-ben-thompson-a_full\"\u003e\n  Stratechery by Ben Thompson (A_full)\n  \u003ca class=\"heading-link\" href=\"#stratechery-by-ben-thompson-a_full\"\u003e\n    \u003ci class=\"fa-solid fa-link\" aria-hidden=\"true\" title=\"Link to heading\"\u003e\u003c/i\u003e\n    \u003cspan class=\"sr-only\"\u003eLink to heading\u003c/span\u003e\n  \u003c/a\u003e\n\u003c/h3\u003e\n\u003cul\u003e\n\u003cli\u003e\u003cstrong\u003e\u003ca href=\"https://stratechery.com/2026/an-interview-with-colossus-eic-jeremy-stern-about-profiling-mark-zuckerberg/\"  class=\"external-link\" target=\"_blank\" rel=\"noopener\"\u003eAn Interview with Colossus EIC Jeremy Stern About Profiling Mark Zuckerberg\u003c/a\u003e\u003c/strong\u003e\n\u003cul\u003e\n\u003cli\u003ePublication Time: 2026-09-24 18:00 Beijing Time\u003c/li\u003e\n\u003cli\u003eAbstract: - An interview with Colossus EIC Jeremy Stern about profiling Mark Zuckerberg and other prominent tech figures.\n\u003cul\u003e\n\u003cli\u003e\u003cstrong\u003e$15\u003c/strong\u003e/month* or *\u003cstrong\u003e$150\u003c/strong\u003e/year.\u003c/li\u003e\n\u003cli\u003eSubstantive analysis of the day\u0026rsquo;s news through three weekly emails or a podcast.\u003c/li\u003e\n\u003cli\u003e\u003cstrong\u003eStrategy interviews\u003c/strong\u003e.\u003c/li\u003e\n\u003cli\u003eInterviews with leading public company CEOs, private company founders, and discussions with fellow analysts.\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003eEN Highlights:\n\u003cul\u003e\n\u003cli\u003eAn interview with Colossus EIC Jeremy Stern about profiling Mark Zuckerberg and other prominent tech figures.\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003c/ul\u003e\n\u003ch3 id=\"google-deepmind-blog-a_full\"\u003e\n  Google DeepMind Blog (A_full)\n  \u003ca class=\"heading-link\" href=\"#google-deepmind-blog-a_full\"\u003e\n    \u003ci class=\"fa-solid fa-link\" aria-hidden=\"true\" title=\"Link to heading\"\u003e\u003c/i\u003e\n    \u003cspan class=\"sr-only\"\u003eLink to heading\u003c/span\u003e\n  \u003c/a\u003e\n\u003c/h3\u003e\n\u003cul\u003e\n\u003cli\u003e\u003cstrong\u003e\u003ca href=\"https://deepmind.google/blog/introducing-gemini-38-live-with-live-avatar/\"  class=\"external-link\" target=\"_blank\" rel=\"noopener\"\u003eIntroducing Gemini 3.8 Live with Live Avatar\u003c/a\u003e\u003c/strong\u003e\n\u003cul\u003e\n\u003cli\u003ePublication Time: 2026-09-25 00:20 Beijing Time\u003c/li\u003e\n\u003cli\u003eAbstract: - Introducing Gemini 3.8 Live and Live Avatar.\n\u003cul\u003e\n\u003cli\u003eThis article from the Google DeepMind Blog explains how the introduction of Gemini 3.8 Live with Live Avatar is shaping the broader AI and infrastructure landscape.\u003c/li\u003e\n\u003cli\u003eIt also lays out the practical implications for founders, operators, and investors following the launch of Gemini 3.8 Live with Live Avatar.\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003eEN Highlights:\n\u003cul\u003e\n\u003cli\u003eIntroducing Gemini 3.8 Live with Live Avatar\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003c/ul\u003e\n\u003ch3 id=\"two-minute-papers-b_introsearch\"\u003e\n  Two Minute Papers (B_intro+search)\n  \u003ca class=\"heading-link\" href=\"#two-minute-papers-b_introsearch\"\u003e\n    \u003ci class=\"fa-solid fa-link\" aria-hidden=\"true\" title=\"Link to heading\"\u003e\u003c/i\u003e\n    \u003cspan class=\"sr-only\"\u003eLink to heading\u003c/span\u003e\n  \u003c/a\u003e\n\u003c/h3\u003e\n\u003cul\u003e\n\u003cli\u003e\u003cstrong\u003e\u003ca href=\"https://www.youtube.com/watch?v=SA9kdAX2Zj0\"  class=\"external-link\" target=\"_blank\" rel=\"noopener\"\u003eClaude Opus 5.5 AI: An Incredible Leap Forward\u003c/a\u003e\u003c/strong\u003e\n\u003cul\u003e\n\u003cli\u003ePublication Time: 2026-09-24 16:40 Beijing Time\u003c/li\u003e\n\u003cli\u003eAbstract: - ❤️ Check out Lambda here and sign up for their GPU Cloud:.\n\u003cul\u003e\n\u003cli\u003eNote: In the walking creatures experiment, Astra used a simplified model and was unable to implement the correct one.\u003c/li\u003e\n\u003cli\u003eThings did not improve after simulating it for more generations.\u003c/li\u003e\n\u003cli\u003eThe walking creatures video and paper:.\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003eEN Highlights:\n\u003cul\u003e\n\u003cli\u003e❤️ Check out Lambda here and sign up for their GPU Cloud:\u003c/li\u003e\n\u003cli\u003eNote: in the walking creatures experiment, Astra used a simplified model and was unable to implement the correct one\u003c/li\u003e\n\u003cli\u003eThings did not improve after simulating it for more generations\u003c/li\u003e\n\u003cli\u003eClaude Opus 5.5:\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003c/ul\u003e\n\u003ch3 id=\"arxiv-csai-b_introsearch\"\u003e\n  ArXiv cs.AI (B_intro+search)\n  \u003ca class=\"heading-link\" href=\"#arxiv-csai-b_introsearch\"\u003e\n    \u003ci class=\"fa-solid fa-link\" aria-hidden=\"true\" title=\"Link to heading\"\u003e\u003c/i\u003e\n    \u003cspan class=\"sr-only\"\u003eLink to heading\u003c/span\u003e\n  \u003c/a\u003e\n\u003c/h3\u003e\n\u003cul\u003e\n\u003cli\u003e\n\u003cp\u003e\u003cstrong\u003e\u003ca href=\"https://arxiv.org/abs/2609.25010\"  class=\"external-link\" target=\"_blank\" rel=\"noopener\"\u003eDo Synthetic Personas Predict Real Audience Response? A Sim-to-Real Study Where a No-Persona Baseline Beats Persona-Based Copy Simulation\u003c/a\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003ePublication Time: 2026-09-24 12:00 Beijing Time\u003c/li\u003e\n\u003cli\u003eAbstract: - arXiv:2609.25010v1 Announcement Type: new.\n\u003cul\u003e\n\u003cli\u003eAbstract: Marketers are increasingly using large language models (LLMs) as \u0026ldquo;synthetic personas\u0026rdquo; to predict audience response to a piece of copy before its release, encouraged by evidence of LLMs mimicking human samples on profile-based conditioning.\u003c/li\u003e\n\u003cli\u003eBut does this prediction actually work for real behavior? Does the persona mechanism help?\u003c/li\u003e\n\u003cli\u003eWe present a sim-to-real validity study using the Upworthy Research Archive \u0026ndash; thousands of headline A/B tests on live traffic with click-through rates measured \u0026ndash; as held-out ground truth.\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003eEN Highlights:\n\u003cul\u003e\n\u003cli\u003earXiv:2609.25010v1 Announce Type: new\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003eAbstract: Marketers increasingly use large language models (LLMs) as \u0026ldquo;synthetic personas\u0026rdquo; to predict how an audience will react to a piece of copy before it shi…\u003c/p\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003eBut is that prediction actually valid against real behaviour - and does the persona machinery help\u003c/p\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003eWe present a sim-to-real validity study using the Upworthy Research Archive - thousands of headline A/B tests on shared real traffic, with measured click-throug…\u003c/p\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003e\u003cstrong\u003e\u003ca href=\"https://arxiv.org/abs/2609.25013\"  class=\"external-link\" target=\"_blank\" rel=\"noopener\"\u003eDo Existing Preconditioners Improve Biomedical Tabular Foundation Learning? An Empirical Study on TabPFN Optimization\u003c/a\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003ePublication Time: 2026-09-24 12:00 Beijing Time\u003c/li\u003e\n\u003cli\u003eAbstract: - arXiv:2609.25013v1 Announce Type: new.\n\u003cul\u003e\n\u003cli\u003eAbstract: Tabular foundation models have recently shown strong potential for structured biomedical data analysis.\u003c/li\u003e\n\u003cli\u003eAmong them, TabPFN has emerged as an effective approach for low-data tabular classification tasks.\u003c/li\u003e\n\u003cli\u003eHowever, the impact of optimization and preconditioning strategies on biomedical fine-tuning remains largely unexplored.\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003eEN 要点:\n\u003cul\u003e\n\u003cli\u003earXiv:2609.25013v1 Announce Type: new\u003c/li\u003e\n\u003cli\u003eAbstract: Tabular foundation models have recently shown strong potential for structured biomedical data analysis\u003c/li\u003e\n\u003cli\u003eAmong them, TabPFN has emerged as an effective approach for low-data tabular classification tasks\u003c/li\u003e\n\u003cli\u003eHowever, the impact of optimization and preconditioning strategies on biomedical fine-tuning remains largely unexplored\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003e\u003cstrong\u003e\u003ca href=\"https://arxiv.org/abs/2609.25036\"  class=\"external-link\" target=\"_blank\" rel=\"noopener\"\u003e4DGS-JEPA: Temporally Compositional Joint-Embedding Prediction for Dynamic Gaussian Splatting\u003c/a\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003ePublication Time: 2026-09-24 12:00 Beijing Time\u003c/li\u003e\n\u003cli\u003eAbstract: - arXiv:2609.25036v1 Announce Type: new.\n\u003cul\u003e\n\u003cli\u003eAbstract: Dynamic Gaussian Splatting provides an explicit representation of evolving 3D scenes, but existing approaches are primarily optimized for reconstruction, future state generation, or rendering, rather than for learning reusable predictive dynamics.\u003c/li\u003e\n\u003cli\u003eWe propose 4DGS-JEPA, a Gaussian-native joint-embedding predictive architecture for causal multi-horizon prediction over dynamic Gaussian scenes.\u003c/li\u003e\n\u003cli\u003eThe model uses hierarchical scene, motion group, and Gaussian-level representations along with horizon-conditioned transformation operators that support direct prediction and recursive rollout.\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003eEN 要点:\n\u003cul\u003e\n\u003cli\u003earXiv:2609.25036v1 Announce Type: new\u003c/li\u003e\n\u003cli\u003eAbstract: Dynamic Gaussian Splatting provides an explicit representation of evolving 3D scenes, but existing approaches are primarily optimized for reconstructi…\u003c/li\u003e\n\u003cli\u003eWe propose 4DGS-JEPA, a Gaussian-native joint-embedding predictive architecture for causal multi-horizon prediction over dynamic Gaussian scenes\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003eThe model uses a hierarchical scene-, motion-group-, and Gaussian-level representation together with a horizon-conditioned transition operator that supports bot…\u003c/p\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003e\u003cstrong\u003e\u003ca href=\"https://arxiv.org/abs/2609.25088\"  class=\"external-link\" target=\"_blank\" rel=\"noopener\"\u003eAn Accurate and Interpretable Hyper Graph Neural Network for GBM Survival Prediction\u003c/a\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003ePublished: 2026-09-24 12:00 Beijing Time\u003c/li\u003e\n\u003cli\u003eAbstract: - arXiv:2609.25088v1 Announcement Type: new.\n\u003cul\u003e\n\u003cli\u003eAbstract: Survival prediction for glioblastoma multiforme (GBM) demands models that are both accurate and interpretable, yet existing approaches treat these objectives as competing, where high-performance models sacrifice transparency, and interpretable models accept diminished predictive power.\u003c/li\u003e\n\u003cli\u003eWe argue that this trade-off is not inherent.\u003c/li\u003e\n\u003cli\u003eGraph neural networks offer a structural foundation for extracting interpretable, explainable representations without compromising discriminative ability.\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003eEN Key Points:\n\u003cul\u003e\n\u003cli\u003earXiv:2609.25088v1 Announce Type: new\u003c/li\u003e\n\u003cli\u003eAbstract: Survival prediction for glioblastoma multiforme (GBM) demands models that are both accurate and interpretable, yet existing approaches treat these obj…\u003c/li\u003e\n\u003cli\u003eWe argue that this trade-off is not inherent\u003c/li\u003e\n\u003cli\u003eGraph neural net- works offer a structural foundation for extracting interpretable, explainable representations without compromising discriminative ability\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003e\u003cstrong\u003e\u003ca href=\"https://arxiv.org/abs/2609.25165\"  class=\"external-link\" target=\"_blank\" rel=\"noopener\"\u003eOvis-Embedding: Pushing the Frontiers of Universal Omni-Modal Embeddings\u003c/a\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003ePublished: 2026-09-24 12:00 Beijing Time\u003c/li\u003e\n\u003cli\u003eAbstract: - arXiv:2609.25165v1 Announcement Type: new.\n\u003cul\u003e\n\u003cli\u003eAbstract: In this report, we introduce \\textbf{Ovis-Embedding}, a state-of-the-art omni-modal embedding series built on the native integration of text, image, video and audio.\u003c/li\u003e\n\u003cli\u003eInstead of assembling separate modality towers, Ovis-Embedding uses a shared multimodal backbone to encode different modalities in a common representation space.\u003c/li\u003e\n\u003cli\u003eSpecifically, we have made \\textbf{three key advances}: (1) \\textbf{Native omni-modal initialization}: We adopt a pre-trained Qwen-omni model as the embedding backbone and adapt it through contrastive training with low-rank initialization; (2) \\textbf{Data-centric omni-modal training}: We built an extensive, high-quality corpus covering text, images, video, audio, and interleaved multimodal data.\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003eEN Key Points:\n\u003cul\u003e\n\u003cli\u003earXiv:2609.25165v1 Announce Type: new\u003c/li\u003e\n\u003cli\u003eAbstract: In this report, we introduce \\textbf{Ovis-Embedding}, a state-of-the-art omni-modal embedding family built on native integration of text, image, video…\u003c/li\u003e\n\u003cli\u003eInstead of assembling separate modality towers, Ovis-Embedding uses a shared multimodal backbone to encode different modalities in a common representation space\u003c/li\u003e\n\u003cli\u003eSpecifically, we make \\textbf{three key advances}: (1) \\textbf{native omni-modal initialization}: we adopt a pretrained Qwen-omni model as the embedding backbon…\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003e\u003cstrong\u003e\u003ca href=\"https://arxiv.org/abs/2609.25187\"  class=\"external-link\" target=\"_blank\" rel=\"noopener\"\u003eX-Planner: Event-Structured Task Planning for Embodied Intelligence\u003c/a\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003ePublication Time: 2026-09-24 12:00 Beijing Time\u003c/li\u003e\n\u003cli\u003eAbstract: - arXiv:2609.25187v1 Announce Type: new.\n\u003cul\u003e\n\u003cli\u003eAbstract: Task planning connects high-level instructions and executable behaviors in long-horizon operations, but modern Vision-Language-Action (VLA) systems often implicitly convey this intermediate structure.\u003c/li\u003e\n\u003cli\u003eExisting Chain-of-Thought (CoT) planners also tend to rely on coarse task-level annotations or serialize long reasoning traces token by token.\u003c/li\u003e\n\u003cli\u003eWe introduce X-Planner, a planning front-end that addresses the supervision and representation problems of embodied reasoning.\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003eEN Highlights:\n\u003cul\u003e\n\u003cli\u003earXiv:2609.25187v1 Announce Type: new\u003c/li\u003e\n\u003cli\u003eAbstract: Task planning bridges high-level instructions and executable behavior in long-horizon manipulation, yet modern Vision-Language-Action (VLA) systems of…\u003c/li\u003e\n\u003cli\u003eExisting chain-of-thought (CoT) planners also tend to rely on coarse task-level annotations or serialize long reasoning traces token by token\u003c/li\u003e\n\u003cli\u003eWe present X-Planner, a planning front-end that addresses both the supervision and representation of embodied reasoning\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003e\u003cstrong\u003e\u003ca href=\"https://arxiv.org/abs/2609.25199\"  class=\"external-link\" target=\"_blank\" rel=\"noopener\"\u003eLean Pool: An AI-Maintained Archive of Formalized Mathematics\u003c/a\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003ePublication Time: 2026-09-24 12:00 Beijing Time\u003c/li\u003e\n\u003cli\u003eAbstract: - arXiv:2609.25199v1 Announce Type: new.\n\u003cul\u003e\n\u003cli\u003eAbstract: Lean Pool is a repository of formalized mathematics.\u003c/li\u003e\n\u003cli\u003eIt is grown, maintained, and optimized by AI agents.\u003c/li\u003e\n\u003cli\u003earXiv:2609.25199v1 Announce Type: new Abstract: Lean Pool is a repository of formalized mathematics, grown, maintained, and optimized by AI agents.\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003eEN Highlights:\n\u003cul\u003e\n\u003cli\u003earXiv:2609.25199v1 Announce Type: new\u003c/li\u003e\n\u003cli\u003eAbstract: Lean Pool is a repository of formalized mathematics\u003c/li\u003e\n\u003cli\u003eIt is grown, maintained and optimized by AI agents.\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003e\u003cstrong\u003e\u003ca href=\"https://arxiv.org/abs/2609.25254\"  class=\"external-link\" target=\"_blank\" rel=\"noopener\"\u003eThe AI Neuroscientist: An Interactive Agentic Interface for Neuroimaging Analysis\u003c/a\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003ePublication Time: 2026-09-24 12:00 Beijing Time\u003c/li\u003e\n\u003cli\u003eAbstract: - arXiv:2609.25254v1 Announce Type: new.\n\u003cul\u003e\n\u003cli\u003eAbstract: Analyzing neuroimaging data requires specialized coding and statistical expertise, which limits accessibility for researchers without a computational background.\u003c/li\u003e\n\u003cli\u003eWe present the AI Neuroscientist, a language agent for interactive data exploration.\u003c/li\u003e\n\u003cli\u003eThe system integrates large language models (LLMs) with neuroimaging toolkits to perform quality control, modeling, and visualization.\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003eEN Highlights:\n\u003cul\u003e\n\u003cli\u003earXiv:2609.25254v1 Announce Type: new\u003c/li\u003e\n\u003cli\u003eAbstract: Analyzing neuroimaging data requires specialized coding and statistical expertise, which limits accessibility for researchers without computational ba…\u003c/li\u003e\n\u003cli\u003eWe present the AI Neuroscientist, a language agent for interactive data exploration\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003eThe system integrates a large language model (LLM) with a neuroimaging toolset to perform quality control, modeling, and visualization\u003c/p\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003e\u003cstrong\u003e\u003ca href=\"https://arxiv.org/abs/2609.25272\"  class=\"external-link\" target=\"_blank\" rel=\"noopener\"\u003eMedGate-Fusion: Integrating First-Encounter Semantic Narratives and Physiological Biomarkers for Prospective Stroke Risk Stratification\u003c/a\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003ePublication Time: 2026-09-24 12:00 Beijing Time\u003c/li\u003e\n\u003cli\u003eAbstract: - arXiv:2609.25272v1 Announcement Type: New.\n\u003cul\u003e\n\u003cli\u003eAbstract: Prospective stroke risk stratification in primary care is challenging because early risk signals are distributed across routine biomarkers and unstructured clinical narratives.\u003c/li\u003e\n\u003cli\u003eWe propose MedGate-Fusion, a multi-modal gated architecture that integrates transformer-based embeddings of first-encounter narratives with ten routinely recorded risk markers.\u003c/li\u003e\n\u003cli\u003eWe used electronic medical record data from the Canadian Primary Care Sentinel Surveillance Network (CPCSSN).\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003eEN Highlights:\n\u003cul\u003e\n\u003cli\u003earXiv:2609.25272v1 Announce Type: new\u003c/li\u003e\n\u003cli\u003eAbstract: Prospective stroke risk stratification in primary care is challenging because early risk signals are distributed across routine biomarkers and unstruc…\u003c/li\u003e\n\u003cli\u003eWe propose MedGate-Fusion, a multi-modal gated architecture that integrates transformer-based embeddings of first-encounter narratives with ten routinely record…\u003c/li\u003e\n\u003cli\u003eWe used electronic medical record data from the Canadian Primary Care Sentinel Surveillance Network (CPCSSN)\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003e\u003cstrong\u003e\u003ca href=\"https://arxiv.org/abs/2609.25284\"  class=\"external-link\" target=\"_blank\" rel=\"noopener\"\u003eWhen LLM Agents Fail to Read the Room: ReAdapt for Relational Social Reasoning\u003c/a\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003ePublication Time: 2026-09-24 12:00 Beijing Time\u003c/li\u003e\n\u003cli\u003eAbstract: - arXiv:2609.25284v1 Announcement Type: New.\n\u003cul\u003e\n\u003cli\u003eAbstract: A social agent\u0026rsquo;s most basic decisions (Should I react to this post?\u003c/li\u003e\n\u003cli\u003eWho should I contact?) are not purely a matter of content.\u003c/li\u003e\n\u003cli\u003eThe correct action often depends on the underlying relationships between people—strength of connection, reciprocity, interconnectedness—rather than which content is most prominent.\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003eEN Highlights:\n\u003cul\u003e\n\u003cli\u003earXiv:2609.25284v1 Announce Type: new\u003c/li\u003e\n\u003cli\u003eAbstract: A social agent\u0026rsquo;s most basic decisions (should I react to this post\u003c/li\u003e\n\u003cli\u003ewho should I reach out to\u003c/li\u003e\n\u003cli\u003e) are not purely content problems\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003c/ul\u003e\n\u003ch3 id=\"arxiv-cscl-b_introsearch\"\u003e\n  ArXiv cs.CL (B_intro+search)\n  \u003ca class=\"heading-link\" href=\"#arxiv-cscl-b_introsearch\"\u003e\n    \u003ci class=\"fa-solid fa-link\" aria-hidden=\"true\" title=\"Link to heading\"\u003e\u003c/i\u003e\n    \u003cspan class=\"sr-only\"\u003eLink to heading\u003c/span\u003e\n  \u003c/a\u003e\n\u003c/h3\u003e\n\u003cul\u003e\n\u003cli\u003e\n\u003cp\u003e\u003cstrong\u003e\u003ca href=\"https://arxiv.org/abs/2609.26913\"  class=\"external-link\" target=\"_blank\" rel=\"noopener\"\u003eCOMED: The Missing Middle Between Routing and Collaboration in Multi-LLM Inference\u003c/a\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003ePublication Time: 2026-09-24 12:00 Beijing Time\u003c/li\u003e\n\u003cli\u003eAbstract: - arXiv:2609.26913v1 Announcement Type: New.\n\u003cul\u003e\n\u003cli\u003eAbstract: No single large language model (LLM) is consistently reliable across all queries, which has motivated multi-model inference systems to either route between models or combine their outputs.\u003c/li\u003e\n\u003cli\u003eHowever, routing stops after selecting the initial model, while dense collaboration invokes peers on every query.\u003c/li\u003e\n\u003cli\u003eWe demonstrate that collaboration is non-monotonic: peers can recover from failures that a model would have made alone, but they can also corrupt an initially correct answer.\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003eEN Highlights:\n\u003cul\u003e\n\u003cli\u003earXiv:2609.26913v1 Announce Type: new\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003eAbstract: No single Large Language Model (LLM) is uniformly reliable across queries, motivating multi-model inference systems that either route among models or…\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003eHowever, routing stops after selecting an initial model, while dense collaboration invokes peers on every query\u003c/li\u003e\n\u003cli\u003eWe show that collaboration is non-monotonic: peers can recover failures that no model solves alone, but can also corrupt initially correct answers\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003e\u003cstrong\u003e\u003ca href=\"https://arxiv.org/abs/2609.26926\"  class=\"external-link\" target=\"_blank\" rel=\"noopener\"\u003eExperts Rise Where LLMs Disagree: Using Cross-Model Disagreement to Target Expert Effort in LLM Codebook Revision for Large-Scale Annotation\u003c/a\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003ePublication Time: 2026-09-24 12:00 Beijing Time\u003c/li\u003e\n\u003cli\u003eAbstract: - arXiv:2609.26926v1 Announcement Type: new.\n\u003cul\u003e\n\u003cli\u003eAbstract: Large-scale text annotation often brings expert insights to millions of documents through a codebook followed by AI annotators.\u003c/li\u003e\n\u003cli\u003eHowever, developing a robust codebook takes several months.\u003c/li\u003e\n\u003cli\u003eLarge Language Models (LLMs) can accelerate this process by applying an early codebook to the data, surfacing cases with strong LLM disagreement, and soliciting expert feedback to resolve them.\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003eEN Highlights:\n\u003cul\u003e\n\u003cli\u003earXiv:2609.26926v1 Announce Type: new\u003c/li\u003e\n\u003cli\u003eAbstract: Large-scale text annotation brings expert insight to millions of documents, often through a codebook that AI annotators follow\u003c/li\u003e\n\u003cli\u003eDeveloping a robust codebook, however, takes months\u003c/li\u003e\n\u003cli\u003eLarge language models (LLMs) could speed this process by applying an early codebook to the data, surfacing cases with strong LLM disagreement, and eliciting exp…\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003e\u003cstrong\u003e\u003ca href=\"https://arxiv.org/abs/2609.26942\"  class=\"external-link\" target=\"_blank\" rel=\"noopener\"\u003eRecognized but Not Produced: A Generation Benchmark for Culturally Specific Kinship Terms\u003c/a\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003ePublication Time: 2026-09-24 12:00 Beijing Time\u003c/li\u003e\n\u003cli\u003eAbstract: - arXiv:2609.26942v1 Announcement Type: new.\n\u003cul\u003e\n\u003cli\u003eAbstract: Current literature evaluates Large Language Models (LLMs) on multilingual kinship understanding using multiple-choice benchmarks, treating it as a recognition problem.\u003c/li\u003e\n\u003cli\u003eIn contrast, we prompt five open-weight LLMs to generate kinship terms in two communicative tasks across three non-Western languages (Hindi, Tamil, and Korean), and pair this with a matched-options-supported choice baseline.\u003c/li\u003e\n\u003cli\u003eOn the same relational-linguistic units, GPT OSS120B selected the correct term in 90.67% of 75 valid units but generated an acceptable term in 36.00% of corresponding attempts; Llama 3.370B showed the same pattern (77.92% vs 24.24%).\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003eEN Highlights:\n\u003cul\u003e\n\u003cli\u003earXiv:2609.26942v1 Announce Type: new\u003c/li\u003e\n\u003cli\u003eAbstract: Current literature evaluates large language models (LLMs) on multilingual kinship understanding using multiple choice benchmarks, treating it as a rec…\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003eWe instead prompt five open weight LLMs to generate kinship terms in three non Western languages (Hindi, Tamil, and Korean) across two communicative tasks and p…\u003c/p\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003eOn identical relation language cells, GPT OSS120B selects the correct term in 90.67% of 75 valid cells but produces an accepted term in 36.00% of the correspond…\u003c/p\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003e\u003cstrong\u003e\u003ca href=\"https://arxiv.org/abs/2609.26945\"  class=\"external-link\" target=\"_blank\" rel=\"noopener\"\u003eClassifying Interpretive Canons at the Sentence Level: A Benchmark from the German Federal Constitutional Court\u003c/a\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003ePosted: 2026-09-24 12:00 Beijing Time\u003c/li\u003e\n\u003cli\u003eAbstract: - arXiv:2609.26945v1 Announcement Type: new.\n\u003cul\u003e\n\u003cli\u003eAbstract: Judicial reasoning remains challenging for large language models (LLMs) to analyze.\u003c/li\u003e\n\u003cli\u003eThis paper provides a sentence-level benchmark for evaluating the ability of LLMs to classify interpretive canons, as articulated by Larenz in the tradition of Savigny.\u003c/li\u003e\n\u003cli\u003eOur contributions are threefold.\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003eEN Key Points:\n\u003cul\u003e\n\u003cli\u003earXiv:2609.26945v1 Announce Type: new\u003c/li\u003e\n\u003cli\u003eAbstract: Judicial reasoning remains challenging for large language models (LLMs) to analyze\u003c/li\u003e\n\u003cli\u003eThis paper contributes a sentence-level benchmark for evaluating the ability of LLMs to classify interpretive canons as articulated by Larenz in the tradition o…\u003c/li\u003e\n\u003cli\u003eOur contributions are threefold\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003e\u003cstrong\u003e\u003ca href=\"https://arxiv.org/abs/2609.26976\"  class=\"external-link\" target=\"_blank\" rel=\"noopener\"\u003eWhen Learned Context Planning Fails to Beat Strong Retrieval: A Controlled Study of Planning, Routing, and Reranking for Long-Context QA\u003c/a\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003ePosted: 2026-09-24 12:00 Beijing Time\u003c/li\u003e\n\u003cli\u003eAbstract: - arXiv:2609.26976v1 Announcement Type: new.\n\u003cul\u003e\n\u003cli\u003eAbstract: Learned context planning selects evidence atoms before an answer model reasons over them.\u003c/li\u003e\n\u003cli\u003eWe test whether this learned selection improves long-context multiple-choice QA after strong retrieval, routing, budgeted-selector, and reranking controls.\u003c/li\u003e\n\u003cli\u003eOur primary diagnostic uses all 503 LongBench-v2 MCQ questions with Qwen2.5-7B-Instruct.\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003eEN Key Points:\n\u003cul\u003e\n\u003cli\u003earXiv:2609.26976v1 Announce Type: new\u003c/li\u003e\n\u003cli\u003eAbstract: Learned context planning selects evidence atoms before an answer model reasons over them\u003c/li\u003e\n\u003cli\u003eWe test whether this learned selection improves long-context multiple-choice QA after strong retrieval, routing, budgeted-selector, and reranking controls\u003c/li\u003e\n\u003cli\u003eOur primary diagnostic uses all 503 LongBench-v2 MCQ questions with Qwen2.5-7B-Instruct\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003e\u003cstrong\u003e\u003ca href=\"https://arxiv.org/abs/2609.27009\"  class=\"external-link\" target=\"_blank\" rel=\"noopener\"\u003eLEGO: Synergizing Expert GraphRAG and Expert Chain-of-Thought for Legal Reasoning\u003c/a\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003ePosted: 2026-09-24 12:00 Beijing Time\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003eAbstract: - arXiv:2609.27009v1 Announce Type: new.\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003eAbstract: Large language models are increasingly applied to high-risk domains such as law, yet complex legal reasoning remains limited by two structural challenges.\u003c/li\u003e\n\u003cli\u003eFirst, existing RAG and GraphRAG methods emphasize lexical or semantic similarity while overlooking normative relations among legal provisions.\u003c/li\u003e\n\u003cli\u003eSecond, vanilla Chain-of-Thought prompting may generate plausible rationales without enforcing the normative structure of legal reasoning.\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003e\u003cstrong\u003e\u003ca href=\"https://arxiv.org/abs/2609.27032\"  class=\"external-link\" target=\"_blank\" rel=\"noopener\"\u003eLexLattice: Multilingual Extractive Summarization via Neural Cellular Automata on Document Hierarchies\u003c/a\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003eRelease Time: 2026-09-24 12:00 Beijing Time\u003c/li\u003e\n\u003cli\u003eAbstract: - arXiv:2609.27032v1 Announce Type: new.\n\u003cul\u003e\n\u003cli\u003eAbstract: Faithfulness is a central concern in legal text summarization, which motivates extractive approaches that select verbatim content traceable to its source.\u003c/li\u003e\n\u003cli\u003eSuch methods typically rank paragraphs or other structural units in isolation, yet give little attention to consolidating evidence that is distributed across, and shares significance between, the document\u0026rsquo;s more distant parts.\u003c/li\u003e\n\u003cli\u003eWe introduce LexLattice, an extractive summarizer that reifies a legal act\u0026rsquo;s hierarchy as a two-dimensional semantic lattice and consolidates over it with a masked two-dimensional neural cellular automaton before selection.\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003e\u003cstrong\u003e\u003ca href=\"https://arxiv.org/abs/2609.27043\"  class=\"external-link\" target=\"_blank\" rel=\"noopener\"\u003eEduBehaviors: Assertion-based Schemas for Auditable Coding of Educational Dialogues\u003c/a\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003eRelease Time: 2026-09-24 12:00 Beijing Time\u003c/li\u003e\n\u003cli\u003eAbstract: - arXiv:2609.27043v1 Announce Type: new.\n\u003cul\u003e\n\u003cli\u003eAbstract: Large language models allow for the rapid deployment of instructional annotations corresponding to structures of interest, allowing natural language interfaces to generate classifications on conversational datasets.\u003c/li\u003e\n\u003cli\u003eHowever, due to the opacity of LLM reasoning, we are unable to have verifiable, mechanistic insight into the reasons a model chooses a label for an utterance.\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003eWe introduce the EduBehaviors framework, an interpretable, scalable approach to annotating educational data that uses LLMs to measure repeated observable behaviors related to many constructs of interest, and then learns classifiers for those constructs based on these observable behaviors.\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003eEN Highlights:\n\u003cul\u003e\n\u003cli\u003earXiv:2609.27043v1 Announce Type: new\u003c/li\u003e\n\u003cli\u003eAbstract: Large language models have allowed the rapid deployment of pedagogical annotations corresponding to constructs of interest, allowing a natural languag…\u003c/li\u003e\n\u003cli\u003eHowever due to the opaque nature of LLM reasoning, we have no verifiable, mechanistic insight into why a model chose a label for an utterance\u003c/li\u003e\n\u003cli\u003eWe introduce the EduBehaviors framework, an interpretable, scalable approach to annotating educational data that uses LLMs to measure repeated observable behavi…\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003e\u003cstrong\u003e\u003ca href=\"https://arxiv.org/abs/2609.27059\"  class=\"external-link\" target=\"_blank\" rel=\"noopener\"\u003eThe Illinois Social Attitudes Aggregate Corpus (ISAAC): An Open Tool and Reproducible Pipeline for Analyzing Social Group Discourse at Scale\u003c/a\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003ePublished: 2026-09-24 12:00 Beijing Time\u003c/li\u003e\n\u003cli\u003eAbstract: - arXiv:2609.27059v1 Announce Type: new.\n\u003cul\u003e\n\u003cli\u003eAbstract: We introduce the Illinois Social Attitudes Aggregate Corpus (ISAAC), an open, modular, and accessible corpus of over 527 million English Reddit posts related to six key social group characteristics based on factors such as race, gender, age, ability, weight, and skin color, covering a 17-year period from 2007 to 2023.\u003c/li\u003e\n\u003cli\u003eA multi-step, human-audited filtering pipeline was used to keep irrelevant content in the curated dataset below 10%, both overall and for each distinct social group.\u003c/li\u003e\n\u003cli\u003eEach post was then algorithmically annotated with the user\u0026rsquo;s estimated home region, along with a suite of validated off-the-shelf and custom semantic labels, including morality, emotion, sentiment, and linguistic generalizations.\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003eEN Highlights:\n\u003cul\u003e\n\u003cli\u003earXiv:2609.27059v1 Announce Type: new\u003c/li\u003e\n\u003cli\u003eAbstract: We introduce the Illinois Social Attitudes Aggregate Corpus (ISAAC), an open, modular, and accessible corpus of 527 million+ English-language Reddit p…\u003c/li\u003e\n\u003cli\u003eA multi-step, human-audited filtering pipeline was used to keep irrelevant content in the curated dataset below 10%, both overall and for each social group dist…\u003c/li\u003e\n\u003cli\u003eEach post was then algorithmically annotated with the user\u0026rsquo;s estimated home region, along with a suite of validated off-the-shelf and custom semantic labels inc…\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003e\u003cstrong\u003e\u003ca href=\"https://arxiv.org/abs/2609.27064\"  class=\"external-link\" target=\"_blank\" rel=\"noopener\"\u003eWhat Changes When Fact-Verification Scores Improve? Evidence and Answer Accounting Across Trained Verifiers and LLMs\u003c/a\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003ePublished: 2026-09-24 12:00 Beijing Time\u003c/li\u003e\n\u003cli\u003eAbstract: - arXiv:2609.27064v1 Announce Type: new.\n\u003cul\u003e\n\u003cli\u003eAbstract: Joint fact-verification scores evaluate answers and submitted evidence together.\u003c/li\u003e\n\u003cli\u003eWhen scores improve, if the answer remains unchanged, how much gain is still left?\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003eOn FEVEROUS, the strict score is the percentage of claims with a correct answer and a complete annotated evidence group in the submitted evidence.\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003eEN Highlights:\n\u003cul\u003e\n\u003cli\u003earXiv:2609.27064v1 Announce Type: new\u003c/li\u003e\n\u003cli\u003eAbstract: A joint fact-verification score assesses answers and submitted evidence together\u003c/li\u003e\n\u003cli\u003eWhen the score improves, how much of the gain remains if the answers are held fixed\u003c/li\u003e\n\u003cli\u003eOn FEVEROUS, strict score is the percentage of claims with a correct answer and a complete annotated evidence group in the submitted evidence\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003c/ul\u003e\n\u003ch3 id=\"arxiv-cslg-b_introsearch\"\u003e\n  ArXiv cs.LG (B_intro+search)\n  \u003ca class=\"heading-link\" href=\"#arxiv-cslg-b_introsearch\"\u003e\n    \u003ci class=\"fa-solid fa-link\" aria-hidden=\"true\" title=\"Link to heading\"\u003e\u003c/i\u003e\n    \u003cspan class=\"sr-only\"\u003eLink to heading\u003c/span\u003e\n  \u003c/a\u003e\n\u003c/h3\u003e\n\u003cul\u003e\n\u003cli\u003e\n\u003cp\u003e\u003cstrong\u003e\u003ca href=\"https://arxiv.org/abs/2609.26811\"  class=\"external-link\" target=\"_blank\" rel=\"noopener\"\u003eThe Drift Contract: Spectral Updates for Depth-Robust Local Learning\u003c/a\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003ePublication Time: 2026-09-24 12:00 Beijing Time\u003c/li\u003e\n\u003cli\u003eAbstract: - arXiv:2609.26811v1 Announce Type: new.\n\u003cul\u003e\n\u003cli\u003eAbstract: Local learning trains each layer with its own auxiliary loss and no global backward pass, which makes layer updates structurally parallel.\u003c/li\u003e\n\u003cli\u003eTwo problems have kept it marginal: accuracy degrades as depth grows, and hyperparameters are fragile.\u003c/li\u003e\n\u003cli\u003eWe apply Muon-style spectral update geometry (momentum orthogonalization with spectral step scaling) to per-layer local updates, an intersection not previously studied.\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003eEN Highlights:\n\u003cul\u003e\n\u003cli\u003earXiv:2609.26811v1 Announce Type: new\u003c/li\u003e\n\u003cli\u003eAbstract: Local learning trains each layer with its own auxiliary loss and no global backward pass, which makes layer updates structurally parallel\u003c/li\u003e\n\u003cli\u003eTwo problems have kept it marginal: accuracy degrades as depth grows, and hyperparameters are fragile\u003c/li\u003e\n\u003cli\u003eWe apply Muon-style spectral update geometry (momentum orthogonalization with spectral step scaling) to per-layer local updates, an intersection not previously…\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003e\u003cstrong\u003e\u003ca href=\"https://arxiv.org/abs/2609.26820\"  class=\"external-link\" target=\"_blank\" rel=\"noopener\"\u003eSignal2Symbol: Neuro-Symbolic Temporal Reasoning for Explainable Physiological Time-Series Anomaly Detection\u003c/a\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003ePublication Time: 2026-09-24 12:00 Beijing Time\u003c/li\u003e\n\u003cli\u003eAbstract: - arXiv:2609.26820v1 Announce Type: new.\n\u003cul\u003e\n\u003cli\u003eAbstract: Physiological time series such as electrocardiograms (ECG) and electroencephalograms (EEG) exhibit complex temporal structure, substantial acquisition variability, and a strong need for transparent decision-making.\u003c/li\u003e\n\u003cli\u003eAlthough deep models can achieve high detection performance, they often provide only limited insight into why a segment is anomalous, how local anomalies evolve over time, and whether a detection is part of a broader, recurring pattern.\u003c/li\u003e\n\u003cli\u003eWe propose Signal2Symbol, a neuro-symbolic framework for explainable biosignal anomaly detection.\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003eEN Highlights:\n\u003cul\u003e\n\u003cli\u003earXiv:2609.26820v1 Announce Type: new\u003c/li\u003e\n\u003cli\u003eAbstract: Physiological time series such as electrocardiograms (ECG) and electroencephalograms (EEG) exhibit complex temporal structure, substantial acquisition…\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003eAlthough deep models can achieve high detection performance, they often provide limited insight into why a segment is anomalous, how local anomalies relate over…\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003eWe propose Signal2Symbol, a neuro-symbolic framework for explainable biosignal anomaly detection\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003e\u003cstrong\u003e\u003ca href=\"https://arxiv.org/abs/2609.26822\"  class=\"external-link\" target=\"_blank\" rel=\"noopener\"\u003eHARN: Hierarchical Associative Resonance Network for Event-Driven Multi-Timeframe Forecasting\u003c/a\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003ePublished: 2026-09-24 12:00 Beijing Time\u003c/li\u003e\n\u003cli\u003eAbstract: - arXiv:2609.26822v1 Announcement Type: new.\n\u003cul\u003e\n\u003cli\u003eAbstract: Financial time series evolve across multiple temporal resolutions, posing a challenge for forecasting systems to incorporate newly available information without repetitive recalculation of unchanged representations.\u003c/li\u003e\n\u003cli\u003eWe introduce HARN, a Hierarchical Associative Resonance Network for event-driven multi-timeframe forecasting.\u003c/li\u003e\n\u003cli\u003eHARN maintains persistent representations across temporal levels and updates each level only when its corresponding complete bar becomes available.\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003eEN Highlights:\n\u003cul\u003e\n\u003cli\u003earXiv:2609.26822v1 Announce Type: new\u003c/li\u003e\n\u003cli\u003eAbstract: Financial time series evolve across multiple temporal resolutions, challenging forecasting systems to incorporate newly available information without…\u003c/li\u003e\n\u003cli\u003eWe introduce HARN, a Hierarchical Associative Resonance Network for event-driven multi-timeframe forecasting\u003c/li\u003e\n\u003cli\u003eHARN maintains persistent representations across temporal levels and updates each level only when its corresponding completed bar becomes available\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003e\u003cstrong\u003e\u003ca href=\"https://arxiv.org/abs/2609.26826\"  class=\"external-link\" target=\"_blank\" rel=\"noopener\"\u003eWhat Makes a Terminal-Bench Task Hard? Separating Genuine Hardness from Fake-Hardness on an Adjudicated Agentic Corpus\u003c/a\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003ePublished: 2026-09-24 12:00 Beijing Time\u003c/li\u003e\n\u003cli\u003eAbstract: - arXiv:2609.26826v1 Announcement Type: new.\n\u003cul\u003e\n\u003cli\u003eAbstract: Frontier benchmarks require tasks that current models cannot solve.\u003c/li\u003e\n\u003cli\u003eHowever, a task that no model can solve is not necessarily a hard task.\u003c/li\u003e\n\u003cli\u003eThe same zero pass rate can stem from an actual capability gap, but it can also come from missing context, a broken reference solution, infrastructure failure, or a bypassable validator.\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003eEN Highlights:\n\u003cul\u003e\n\u003cli\u003earXiv:2609.26826v1 Announce Type: new\u003c/li\u003e\n\u003cli\u003eAbstract: Frontier benchmarks need tasks that current models cannot solve\u003c/li\u003e\n\u003cli\u003eBut a task that no model solves is not automatically a hard task\u003c/li\u003e\n\u003cli\u003eThe same zero pass rate can come from a real capability gap, but it can also come from missing context, a broken reference solution, infrastructure failure, or…\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003e\u003cstrong\u003e\u003ca href=\"https://arxiv.org/abs/2609.26839\"  class=\"external-link\" target=\"_blank\" rel=\"noopener\"\u003eLWCal: Loss-Weighted Calibration for Tabular Classifiers with Noisy Calibration Labels\u003c/a\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003ePublished: 2026-09-24 12:00 Beijing Time\u003c/li\u003e\n\u003cli\u003eAbstract: - arXiv:2609.26839v1 Announcement Type: new.\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003eAbstract: Post-hoc probability calibration is usually evaluated under an optimistic assumption: the held-out calibration labels are clean.\u003c/p\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003eIn many AI deployment settings, however, labels come from weak annotators, historical decisions, heuristics, or distant supervision, so the same label noise that corrupts training also corrupts calibration.\u003c/p\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003eWe study this overlooked failure mode for tabular classifiers and propose LWCal, a CPU-only post-hoc calibrator that down-weights calibration examples whose noisy labels contradict the base model\u0026rsquo;s held-out probabilities.\u003c/p\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003eEN Highlights:\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003earXiv:2609.26839v1 Announce Type: new\u003c/li\u003e\n\u003cli\u003eAbstract: Post-hoc probability calibration is usually evaluated under an optimistic assumption: the held-out calibration labels are clean\u003c/li\u003e\n\u003cli\u003eIn many AI deployment settings, however, labels come from weak annotators, historical decisions, heuristics, or distant supervision, so the same label noise tha…\u003c/li\u003e\n\u003cli\u003eWe study this overlooked failure mode for tabular classifiers and propose LWCal, a CPU-only post-hoc calibrator that down-weights calibration examples whose noi…\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003e\u003cstrong\u003e\u003ca href=\"https://arxiv.org/abs/2609.26848\"  class=\"external-link\" target=\"_blank\" rel=\"noopener\"\u003eA Leakage-Aware Multimodal Evaluation Framework for Early Intraoperative Acute Kidney Injury Prediction\u003c/a\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003ePublication Time: 2026-09-24 12:00 Beijing Time\u003c/li\u003e\n\u003cli\u003eAbstract: - arXiv:2609.26848v1 Announce Type: new.\n\u003cul\u003e\n\u003cli\u003eAbstract: The incidence of postoperative acute kidney injury (AKI) after major non-cardiac surgery is high, but early intraoperative risk stratification remains difficult.\u003c/li\u003e\n\u003cli\u003eIn this retrospective cohort study, we propose SynerT, a waveform-only hybrid temporal backbone that combines a causal dilated TCN with a hierarchy of dilated recurrent layers to encode early intraoperative physiological trajectories for AKI risk prediction.\u003c/li\u003e\n\u003cli\u003eBuilding on SynerT, we further design two model variants that extend the backbone with structured clinical context: SynerT-MM, a late-fusion multimodal extension integrating hemodynamic load summaries and preoperative covariates; and SynerTStack, a leakage-safe stacked ensemble that combines SynerT-MM\u0026rsquo;s cross-validated predictions with a strong tabular baseline in a meta-learning phase.\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003eEN Highlights:\n\u003cul\u003e\n\u003cli\u003earXiv:2609.26848v1 Announce Type: new\u003c/li\u003e\n\u003cli\u003eAbstract: Postoperative acute kidney injury (AKI) after major non-cardiac surgery carries substantial morbidity, yet early intraoperative risk stratification re…\u003c/li\u003e\n\u003cli\u003eIn this retrospective cohort study, we propose SynerT, a waveform-only hybrid temporal backbone that combines a causal dilated TCN with a hierarchy of dilated r…\u003c/li\u003e\n\u003cli\u003eBuilding on SynerT, we further design two model variants that extend the backbone with structured clinical context: SynerT-MM, a late-fusion multimodal extensio…\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003e\u003cstrong\u003e\u003ca href=\"https://arxiv.org/abs/2609.26853\"  class=\"external-link\" target=\"_blank\" rel=\"noopener\"\u003eCOPE: Continual Personalization of LLMs under Sparse User Feedback via User Embeddings and Self-Evaluation\u003c/a\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003ePublication Time: 2026-09-24 12:00 Beijing Time\u003c/li\u003e\n\u003cli\u003eAbstract: - arXiv:2609.26853v1 Announce Type: new.\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003eAbstract: While Large Language Models (LLMs) have achieved remarkable results across various benchmarks, their alignment with normative values often leads to homogenized responses that fail to satisfy diverse user preferences.\u003c/p\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003eExisting training-free methods often occupy valuable context windows through prompt engineering, while training-based methods typically remain static post-training, unable to support the continuous optimization required in real-world environments.\u003c/p\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003eTo address these challenges, we propose COPE (Continual Optimization with Personalized embedding and self-Evaluation), a novel optimization framework tailored for real-world-driven interactive settings with sparse user feedback.\u003c/p\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003eEN Highlights:\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003earXiv:2609.26853v1 Announce Type: new\u003c/li\u003e\n\u003cli\u003eAbstract: While Large Language Models (LLMs) have achieved remarkable results across various benchmarks, their alignment with normative values often results in…\u003c/li\u003e\n\u003cli\u003eExisting training-free methods often occupy valuable context windows through prompt engineering, while training-based methods typically remain static post-train…\u003c/li\u003e\n\u003cli\u003eTo address these challenges, we propose COPE (Continual Optimization with Personalized embedding and self-Evaluation), a novel optimization framework tailored f…\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003e\u003cstrong\u003e\u003ca href=\"https://arxiv.org/abs/2609.26855\"  class=\"external-link\" target=\"_blank\" rel=\"noopener\"\u003eQUARTET: Quad-branch cross-Attention and Random-walk Traces for Enhancing Transformers on Relational Graphs\u003c/a\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003ePublication Time: 2026-09-24 12:00 Beijing Time\u003c/li\u003e\n\u003cli\u003eAbstract: - arXiv:2609.26855v1 Announce Type: new.\n\u003cul\u003e\n\u003cli\u003eAbstract: Relational Deep Learning (RDL) models multi-table databases as heterogeneous temporal graphs, and graph transformers currently achieve state-of-the-art performance on benchmarks like RelBench.\u003c/li\u003e\n\u003cli\u003eHowever, the current leading model, RelGT, suffers from two key limitations: its random local sampler yields loosely connected subgraphs that hinder message passing, and its global attention module relies on a single, seed-feature-based memory, ignoring broader macro-dynamics.\u003c/li\u003e\n\u003cli\u003eTo overcome these limitations, we introduce QUARTET, an expressive graph transformer architecture that applies full self-attention on local subgraphs while enriching the global context via cross-attention branches.\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003eEN Highlights:\n\u003cul\u003e\n\u003cli\u003earXiv:2609.26855v1 Announce Type: new\u003c/li\u003e\n\u003cli\u003eAbstract: Relational Deep Learning (RDL) models multi-table databases as heterogeneous temporal graphs, and graph transformers currently achieve state-of-the-ar…\u003c/li\u003e\n\u003cli\u003eHowever, the current leading model, RelGT, suffers from two key limitations: its random local sampler yields loosely connected subgraphs that hinder message pas…\u003c/li\u003e\n\u003cli\u003eTo overcome these limitations, we introduce QUARTET, an expressive graph transformer architecture that applies full self-attention on local subgraphs while enri…\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003e\u003cstrong\u003e\u003ca href=\"https://arxiv.org/abs/2609.26866\"  class=\"external-link\" target=\"_blank\" rel=\"noopener\"\u003eMarginally Correct Tool Caches Can Reverse Group-Normalized Policy Updates\u003c/a\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003ePublication Time: 2026-09-24 12:00 Beijing Time\u003c/li\u003e\n\u003cli\u003eAbstract: - arXiv:2609.26866v1 Announce Type: new.\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003eAbstract: Tool-result caching reduces repeated execution in agent training, but also couples rollout randomness.\u003c/p\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003eWe study a two-action model in which independent and shared execution preserve every rollout\u0026rsquo;s conditional reward distribution.\u003c/p\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003eDespite this marginal agreement, sharing one stochastic result per group can reverse the expected group-normalized policy update.\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003eEN Key Points:\n\u003cul\u003e\n\u003cli\u003earXiv:2609.26866v1 Announce Type: new\u003c/li\u003e\n\u003cli\u003eAbstract: Tool-result caching reduces repeated execution in agent training, but also couples rollout randomness\u003c/li\u003e\n\u003cli\u003eWe study a two-action model in which independent and shared execution preserve every rollout\u0026rsquo;s conditional reward distribution\u003c/li\u003e\n\u003cli\u003eDespite this marginal agreement, sharing one stochastic result per group can reverse the expected group-normalized policy update\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003e\u003cstrong\u003e\u003ca href=\"https://arxiv.org/abs/2609.26890\"  class=\"external-link\" target=\"_blank\" rel=\"noopener\"\u003ePR-Smoother: Simulator-Preserving Non-Gaussian Smoothing for Data Assimilation\u003c/a\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003ePublished: 2026-09-24 12:00 Beijing Time\u003c/li\u003e\n\u003cli\u003eAbstract: - arXiv:2609.26890v1 Announce Type: new.\n-Abstract: Many physical data assimilation (DA) workflows require smoothing methods that represent non-Gaussian posteriors over physical state variables, scale to high-dimensional simulators, train only from an observation window, and remain compatible with the calibration of a prescribed simulator.\n\u003cul\u003e\n\u003cli\u003eWe introduce PR-Smoother, a simulator-preserving amortized smoother designed for this prescribed-simulator DA regime.\u003c/li\u003e\n\u003cli\u003eIts key design principle is to keep the prescribed simulator explicit in both the evidence lower bound and the variational family: rather than learning replacement dynamics or trajectories, PR-Smoother only learns future conditional corrections around a prescribed rollout.\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003eEN Key Points:\n\u003cul\u003e\n\u003cli\u003earXiv:2609.26890v1 Announce Type: new\u003c/li\u003e\n\u003cli\u003eAbstract: Many physical data assimilation (DA) workflows require smoothing methods that represent non-Gaussian posteriors over physical state variables, scale t…\u003c/li\u003e\n\u003cli\u003eWe introduce PR-Smoother, a simulator-preserving amortized smoother designed for this prescribed-simulator DA regime\u003c/li\u003e\n\u003cli\u003eIts key design principle is to keep the prescribed simulator explicit in both the evidence lower bound and the variational family: rather than learning replacem…\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003c/ul\u003e\n",
  "wordCount": 7134,
  "readingTime": 34,
  "tableOfContents": "\u003cnav id=\"TableOfContents\"\u003e\n  \u003cul\u003e\n    \u003cli\u003e\u003ca href=\"#-this-issues-watch-list-a-deep-dive\"\u003e📖 This Issue\u0026rsquo;s Watch List: A Deep Dive\u003c/a\u003e\u003c/li\u003e\n    \u003cli\u003e\u003ca href=\"#-ai-hot-topics-on-x\"\u003e🌐 AI Hot Topics on X\u003c/a\u003e\n      \u003cul\u003e\n        \u003cli\u003e\u003ca href=\"#topic-1-nous-research-launches-bot-screen-for-live-ai-agent-monitoring\"\u003eTopic 1: Nous Research Launches Bot Screen for Live AI Agent Monitoring\u003c/a\u003e\u003c/li\u003e\n        \u003cli\u003e\u003ca href=\"#topic-2-bitget-crypto-exchange-hit-by-3516-million-hot-wallet-breach\"\u003eTopic 2: Bitget Crypto Exchange Hit by $351.6 Million Hot Wallet Breach\u003c/a\u003e\u003c/li\u003e\n        \u003cli\u003e\u003ca href=\"#topic-3-house-report-accuses-doj-of-secretly-targeting-republican-lawmakers\"\u003eTopic 3: House Report Accuses DOJ of Secretly Targeting Republican Lawmakers\u003c/a\u003e\u003c/li\u003e\n        \u003cli\u003e\u003ca href=\"#topic-4-alexandr-wangs-camo-crocs-outfit-sparks-luxury-memes-at-meta-connect\"\u003eTopic 4: Alexandr Wang\u0026rsquo;s Camo Crocs Outfit Sparks Luxury Memes at Meta Connect\u003c/a\u003e\u003c/li\u003e\n        \u003cli\u003e\u003ca href=\"#topic-5-clm-model-speeds-up-ai-agent-decisions-with-contrastive-learning\"\u003eTopic 5: CLM Model Speeds Up AI Agent Decisions with Contrastive Learning\u003c/a\u003e\u003c/li\u003e\n        \u003cli\u003e\u003ca href=\"#topic-6-anthropic-launches-cloud-sessions-for-claude-code-with-subscriber-credits\"\u003eTopic 6: Anthropic Launches Cloud Sessions for Claude Code with Subscriber Credits\u003c/a\u003e\u003c/li\u003e\n        \u003cli\u003e\u003ca href=\"#topic-7-ai-agent-trims-400000-lines-of-redundant-tests-in-openclaw-project\"\u003eTopic 7: AI Agent Trims 400,000 Lines of Redundant Tests in OpenClaw Project\u003c/a\u003e\u003c/li\u003e\n        \u003cli\u003e\u003ca href=\"#topic-8-paper-rex-beats-team-liquid-2-1-to-top-valorant-champions-odds\"\u003eTopic 8: Paper Rex Beats Team Liquid 2-1 to Top VALORANT Champions Odds\u003c/a\u003e\u003c/li\u003e\n        \u003cli\u003e\u003ca href=\"#topic-9-israeli-forward-scores-then-sent-off-for-flag-celebration\"\u003eTopic 9: Israeli Forward Scores Then Sent Off for Flag Celebration\u003c/a\u003e\u003c/li\u003e\n        \u003cli\u003e\u003ca href=\"#topic-10-dagestani-fighters-relentless-discipline-powers-ufc-dominance\"\u003eTopic 10: Dagestani Fighters\u0026rsquo; Relentless Discipline Powers UFC Dominance\u003c/a\u003e\u003c/li\u003e\n        \u003cli\u003e\u003ca href=\"#topic-11-van-dijk-commits-to-netherlands-as-gakpos-late-goal-earns-draw-vs-klopps-germany\"\u003eTopic 11: Van Dijk Commits to Netherlands as Gakpo\u0026rsquo;s Late Goal Earns Draw vs. Klopp\u0026rsquo;s Germany\u003c/a\u003e\u003c/li\u003e\n        \u003cli\u003e\u003ca href=\"#topic-12-lewis-hamilton-shares-ai-video-of-f1-drivers-in-funny-car-ride\"\u003eTopic 12: Lewis Hamilton Shares AI Video of F1 Drivers in Funny Car Ride\u003c/a\u003e\u003c/li\u003e\n        \u003cli\u003e\u003ca href=\"#topic-13-chiles-golden-retriever-puppies-nap-after-military-parade\"\u003eTopic 13: Chile\u0026rsquo;s Golden Retriever Puppies Nap After Military Parade\u003c/a\u003e\u003c/li\u003e\n        \u003cli\u003e\u003ca href=\"#topic-14-birdman-and-toni-braxton-spotted-together-after-rocky-marriage-timeline\"\u003eTopic 14: Birdman and Toni Braxton Spotted Together After Rocky Marriage Timeline\u003c/a\u003e\u003c/li\u003e\n        \u003cli\u003e\u003ca href=\"#topic-15-onlyfans-faces-claims-of-fake-subscribers-and-hidden-motives\"\u003eTopic 15: OnlyFans Faces Claims of Fake Subscribers and Hidden Motives\u003c/a\u003e\u003c/li\u003e\n      \u003c/ul\u003e\n    \u003c/li\u003e\n    \u003cli\u003e\u003ca href=\"#-influencer-insights\"\u003e💡 Influencer Insights\u003c/a\u003e\u003c/li\u003e\n    \u003cli\u003e\u003ca href=\"#-appendix-todays-watch-list-update-source-list\"\u003e📚 Appendix: Today\u0026rsquo;s Watch List Update Source List\u003c/a\u003e\n      \u003cul\u003e\n        \u003cli\u003e\u003ca href=\"#stratechery-by-ben-thompson-a_full\"\u003eStratechery by Ben Thompson (A_full)\u003c/a\u003e\u003c/li\u003e\n        \u003cli\u003e\u003ca href=\"#google-deepmind-blog-a_full\"\u003eGoogle DeepMind Blog (A_full)\u003c/a\u003e\u003c/li\u003e\n        \u003cli\u003e\u003ca href=\"#two-minute-papers-b_introsearch\"\u003eTwo Minute Papers (B_intro+search)\u003c/a\u003e\u003c/li\u003e\n        \u003cli\u003e\u003ca href=\"#arxiv-csai-b_introsearch\"\u003eArXiv cs.AI (B_intro+search)\u003c/a\u003e\u003c/li\u003e\n        \u003cli\u003e\u003ca href=\"#arxiv-cscl-b_introsearch\"\u003eArXiv cs.CL (B_intro+search)\u003c/a\u003e\u003c/li\u003e\n        \u003cli\u003e\u003ca href=\"#arxiv-cslg-b_introsearch\"\u003eArXiv cs.LG (B_intro+search)\u003c/a\u003e\u003c/li\u003e\n      \u003c/ul\u003e\n    \u003c/li\u003e\n  \u003c/ul\u003e\n\u003c/nav\u003e",
  "isDraft": false
}