AI各社の公式発表を1画面に。各社のページを個別に開かなくても最新の動きが追えます。
EgoTools：実世界の一人称動画におけるツール中心の推論
Real-world embodied tasks, from everyday activities to professional procedures, require agents to act under physical constraints while tracking evolving object and task states. Tool use sits at the heart of such tasks, as many everyday and professional activities are tool-mediated. Understanding the…
現在のシーン以降：能動的視点選択によるイベント参照グラスピング
A robot that observes people interacting with objects should be able to carry out later requests that refer back to those interactions. Such requests may specify a grasp target by the role it played in a past event rather than by its name or appearance. Moreover, the target may no longer be visible…
スパース報酬を活かす：マルチ報酬強化学習での密度認識報酬集約
Multi-reward reinforcement learning trains large language models to satisfy multiple behavioral objectives simultaneously. Reward-wise normalization, as used in GDPO, preserves reward-specific relative information within rollout groups, but different objectives can still exhibit uneven learning prog…
信頼性あるAIレビュアーのための欠落要素：レトリカルロバストネスのベンチマーキングからSciCoreレビューへ
AI reviewers can assign different judgments to manuscripts that report the same science in different wording, potentially rewarding rhetorical optimization over scientific improvement. We formulate Rhetorical Robustness as the joint requirement of stability across content-preserving rewrites and dis…
DyRAD: 動的運転シーンのレーダー新規ビュー合成
Reconstructing dynamic driving scenes from recorded sensor data supports closed-loop evaluation of autonomous driving systems by synthesizing observations beyond the original trajectory. Unlike cameras and LiDAR, radar measures radial velocity directly through Doppler. Yet existing radar novel-view…
Agent Error Dataset：失敗分析と誤り対応ファインチューニング向けに50,000の誤り診断ペアをスケーリング
An unsuccessful LLM agent rollout contains more information than its final reward: the observations available to the agent, the actions it chose, and the environment's responses. Reusing this experience for learning requires identifying a decision to revise and testing a concrete alternative. We int…
More Choices, Fewer Decisions：JEVライク直接決定モデルにおける序列スケール偏向
Direct-decision models turn text into low-latency structured labels and scores, making them attractive for classification and automatic evaluation. Yet reliability requires more than accuracy: a model must also use the ordinal decision scale supplied by the user faithfully. We analyze JEV~1.13 and t…
Mid-Harness：ターミナルエージェント向けモデルとハーネス間のアクション規模化
Terminal agents act through stochastic model generations, yet the ability to generate a useful action does not ensure its reliable execution. A poor command (e.g., wrong package install) can change the environment in ways that hinder subsequent progress, even when the model could generate a better a…
EvoDuet：科学的発見に向けたWebサーチとタスク解決の二段階共進化
Evolutionary search with large language models (LLMs) can stall when progress requires external knowledge the model lacks. Supplying relevant documents helps, but simply adding web search tool can keep returning the same pages as solutions change. We introduce EvoDuet, a bi-level optimization method…
False Frontiers：自己進化型検索エージェントの共同不正行為を診断・緩和
Self-evolving search agents build their own training curricula by jointly optimizing a proposer that generates questions and a solver that answers them. This closed loop introduces a failure mode we call co-cheating: the proposer and solver increasingly agree on shared errors, so internal reward imp…
UniEvo-VL：マルチモーダルモデルの自己改善に向けたオンポリシー自己蒸留トレーニング手法
Modern multimodal models bring generation and understanding into a single unified system, which enables them to provide and learn from their own feedback. Motivated by this unified capacity, we introduce UniEvo-VL, a self-evolving framework for multimodal models to learn from this constructive self-…
Open TTS Leaderboard：多言語テキスト音声合成とボイスクローニングのスケーラブル評価
NVIDIA Kumo Tabularが表形式予測の精度と効率で新境地を開く
事実だけでなくソース情報を正確に把握する：MCPエージェント向けソース対応検証
DevDay 2026 総括
Explore more than 20 announcements from OpenAI DevDay 2026, including GPT-6 Astra, ChatGPT, Codex, APIs, security, and new tools for builders.
GPT-6.1 Solを公開
Meet GPT-6.1 Sol: near-Astra intelligence for coding, computer use, and professional work at one-fifth of Astra’s standard API input and output token prices.
FabConおよびSQLCon 2026がバルセロナで開催：Microsoft Copilotとエージェント向けのデータ基盤構築
Microsoft Fabric and SQL innovations announced at FabCon and SQLCon Barcelona 2026 help organizations build trusted data foundations for AI. The post FabCon and SQLCon 2026 in Barcelona: Building the data foundation for Microsoft Copilot and agents appeared first on Microsoft Azure Blog .
LabInstruct：実験室手順のための状況的教育ビデオ生成ベンチマーク
Self-driving laboratories (SDLs) aim to automate the full experimental loop, from scientific decision-making to physical execution. Ideally, AI-generated plans could be carried out directly by robotic systems, but reliable automation remains difficult in complex, open-world laboratory environments,…
In-context Robot Learning Made Simple：マニピュレーションタスク向けの民主化されたレシピ
We study robotic in-context learning (ICL), an emerging paradigm that enables robots to infer and execute tasks from visual demonstrations. Despite its growing promise, the problem itself remains under-defined: a visual demonstration simultaneously conveys action trajectories, object semantics, mani…
STEPQuant：デルタルール再帰状態量子化におけるエラーの時間・位置特性
Linear attention replaces growing KV caches with fixed-size recurrent states, yet these persistent states can become a substantial memory bottleneck under concurrent serving. Directly quantizing recurrent states to low precision often leads to severe accuracy degradation, as quantization errors prop…
ロボット実行エラー下のVLM制御: 自己補償とストレステスト
Vision-language-action (VLA) policies often fail when a robot's executed motion deviates from their commanded action. Such execution errors arise from the robot's mechanics and operating conditions, such as wear and payload changes. We propose self-compensating VLA, a deployment-time adaptation meth…
プロアクティブエージェントの基礎：原則、技術レイヤー、Proactivity-Gym
Proactive LLM agents can turn idle compute into useful support before users ask. Yet even correct work can misread user context, impose review costs, or undermine trust. This work proposes foundations for designing, realizing, and evaluating proactive LLM agents around three joint principles (3T): T…
On-Policy Parameter Update Direction Underlies Generalization in LLM Post-Training
The strong generalization performance of on-policy post-training paradigms has motivated studies of their parameter update behaviors. However, these studies treat the observed behaviors only as byproducts in on-policy training, overlooking their potential to serve as optimization principles for impr…
MotorMind：視覚言語モデルによるロボット操作のゼロショット実装
Vision-language-action (VLA) models have advanced robotic manipulation, but their zero-shot generalization in new tasks and environments remains limited, and their reliance on specialized training keeps them from benefiting directly from rapidly advancing general-purpose vision-language models (VLMs…
異種マルチエージェントLLMのためのプリフィル不要なクロスファミリーKVキャッシュ転送
Recent multi-agent LLM systems increasingly combine heterogeneous models for specialized agent roles. However, text-based communication requires each receiver to prefill shared context already processed by the sender. Reusing the sender's key-value (KV) cache avoids this redundancy, but prefill-free…
Transformerは早期に思考停止する、小規模なLoRAで改善
Pretrained transformers use little of their depth to follow references in context. Thirteen base models reliably follow only 1.4-3.6 lines, and extra pretrained loops add little. A task-trained rank-8 LoRA at one early layer extends this computation with all model weights frozen. Qwen3-8B improves f…
E-MoE: 非ファクタライズド拡散言語モデル向け強化されたMixture-of-Experts
Masked diffusion models (MDMs) generate sequences by progressively unmasking several tokens per denoising step, but their reverse process is typically factorized over positions, limiting sample quality in the few-step regime where diffusion's speed advantage over autoregressive decoding matters most…
TERRA: 筋骨格運動の地形認識再構成・リターゲティング・制御
Recent advances in musculoskeletal modeling and reinforcement learning have enabled muscle-actuated agents to reproduce increasingly complex human motions. Yet these capabilities remain largely confined to flat ground, in part because motion datasets rarely include aligned terrain geometry and becau…
Breaking Babel: 長編字幕翻訳のための自己進化マルチエージェントシステム
Long-form subtitle translation requires reasoning over discourse and cultural context spanning episodes or entire series, while maintaining consistent terminology and style. Existing single-LLM methods are largely sentence-level, and multi-agent systems often use static workflows that do not adapt t…
画像の内容ではなく無関連な文脈がVLMジャッジを不安定にする
Vision-language models (VLMs) are increasingly used in place of human annotators, making it important that substitutability tests reflect the model rather than incidental evaluation conditions. We introduce MIST, the Misleading-Image Stress Test: 200 English sentences, each built around a phrase rea…
EVOKE：転移可能な意思決定に向けた世界知識の引き出し
Large language models (LLMs) are increasingly deployed as agents for multi-step decision-making, yet transfer poorly to unseen environments. World-model methods address this by training agents to predict future observations, at the cost of additional training and errors that compound when prediction…
Learning Meta-Skills for Agent Harness Design in Test-Time AI4AI：テスト時間AI4AIにおけるエージェントハーネス設計のメタスキル学習
Agent performance depends on both reasoning ability and the environment in which it acts. We study test-time AI-for-AI, asking how a Builder can learn to construct better execution environments for a Target while both models' weights remain fixed. To make the Builder's experience reusable, we introd…
AREX-2：長期水平反省的タスクを通じた自己改善エージェントの高度化
We present AREX-2, an effort to advance the self-improving capability of LLM agents, which we define as the ability to iteratively refine a solution at test time. This ability rests on two complementary capabilities: reflection, which produces a solution better than the current one, and long-horizon…
The Teacher Is a Direction, Not a Destination：オンポリシー蒸留におけるRL誘導表現残差の外挿
On-policy distillation (OPD) trains a student to match the teacher's next-token distributions on the student's own trajectories and has yielded substantial empirical gains. Generalized variants allow the student to surpass the teacher by extrapolating an implicit reward in output space. The language…
APM-Bench：一人称ストリーミングビデオアシスタント向けセッション間永続メモリのベンチマーク
To serve as real-world personal assistants, streaming video models need persistent memory that retains past experiences for later use. Yet existing streaming benchmarks and methods often focus on individual continuous videos or short clips, overlooking that real-world interactions are often intermit…
SoL-Refiner：高解像度動画向け光速単一ステップ改善
High-resolution video generation is expensive, as its cost grows rapidly with the number of spatiotemporal tokens. A practical alternative first generates a lower-resolution video and then applies a refiner, but conventional multi-step refinement introduces a second sampling bottleneck. We present S…
EngiWorld：フロンティアエージェントが業務エンジニアリング環境で提供できるもの
Autonomous agents have made rapid progress in general-purpose computer use, but reliable automation of professional industrial engineering remains out of reach, as engineering workflows demand reasoning over geometric and physical constraints and dependencies preserved across software and design sta…
エンティティを追う：エージェント検索向けコーパスマップ
Answering questions and completing tasks over large document collections often requires connecting evidence spread across multiple documents, such as a project's approval recorded in one, its requirements in another, and its latest status in a third. Recent LLM agents approach this by iteratively se…
OpenAI API変更履歴 2026-09-29
Feature Added computer use to the Agents API. Agents can complete tasks in an OpenAI-hosted browser, with website access approvals and sign-in handled by your application. / Feature gpt-6.1-sol v1/responses v1/chat/completions Released GPT-6.1 Sol ( gpt-6.1-sol ) for complex coding and professional…
dotsを導入
Dots by OpenAI are a proactive assistant that can keep working across complex projects and everyday tasks. Learn how dots help you stay in control while work moves forward.
フロンティアAI訓練の安全ケースに向けて
Our early guidelines for safety cases in frontier AI training cover technical safeguards, operational practices, and investigating misalignment incidents
オーストラリアのためにより良い取り組みを
OpenAI apologises for incidents involving Australian government websites and outlines stronger safeguards and support to strengthen Australia’s cyber defences.
Future Vision XPRIZE受賞作『The Gifted』のトレーラーをご覧ください
Gemini 3.8 Flashで構築している4人のビルダーの作品を見る
A collage of four digital 3D visualization projects showing two views of globe satellite maps, a dinosaur skeleton model, and an exploded view of a mechanical transmission diagram with labeled components.
Hallo, Deutschland!
Mistral opens a Munich hub for Physics AI and Industrial AI research, partnering with German industry.
この食品小売業者がGeminiで200人の食事を調理する3つの方法
How Edy’s Grocer Caters for 200 Guests YouTube video
Holo4: 汎用的なコンピュータ利用エージェントの実現
Lenfest InstituteがOpenAI支援拡大で指標的プログラムを成長させる
OpenAI is expanding the Lenfest AI Collaborative and Fellowship Program with $5 million in funding and up to $5 million in software credits and engineering support.
逆蒸留による データ忘却
Multi-step matching models, including flow and diffusion models, produce high-quality outputs but incur substantial inference costs and may reproduce unwanted components of their training datasets. We introduce Inverse Distillation Unlearning (IDU), a unified framework that simultaneously distills a…