毎日更新3935 検証済みのストーリー
AIニュース。 ノイズなしで。
製品の発売、政策の変更、安全性研究、業界の動向についてソースチェックされた AI の報道が、非営利教育チームによって平易な英語で説明されます。
検証済みの調達
すべてのストーリーは、入手可能な最強の証拠、つまり入手可能な場合はオリジナルの情報源、それ以外の場合は明らかに帰属が明記された報道にリンクしています。
平易な英語
何が起こったのか、なぜそれが重要なのか、何を観るべきなのかを、専門用語を使わずに説明します。
フィラーなし
信号が薄い場合は、フィードをパディングするだけで何も公開しません。
さらに多くのストーリー
9 物語革新
Preprint compares how large language models infer opponents’ beliefs in economic games
A new preprint reports that large language models displayed measurable signs of mentalization in two economic games, but their performance and strategies differed by model family and size. The authors say GPT-5 agents adapted to increasingly sophisticated opponents and outperformed the human comparison group in…arxiv.orgセキュリティ
Study finds LLM guardrail approvals can expire before self-adaptive systems act
A paper accepted to ACSOS 2026 reports that language-model guardrails can issue approvals that become invalid before a self-adaptive system uses them. Its proposed Freshness-Bounded Shield reduced measured approval-expiry rates in five simulator environments, though the source does not establish real-world…arxiv.org革新
Benchmark finds local AI agents can handle many hardware-design tool calls, but reliability varies
A new arXiv benchmark finds that open-source AI agents can complete many dependency-ordered hardware-design operations through MCP tools, while tool descriptions, context length and agent configuration strongly affect reliability.arxiv.org革新
SKILL.state proposes explicit execution state for longer-running AI agents
A paper accepted at EMNLP presents a runtime architecture that replaces growing agent conversation histories with mutable structured execution state.arxiv.org革新
FinRiskAtlas finds broad AI scores can miss weaknesses in financial risk review
A new Chinese-language benchmark evaluates large language models by the decisions and evidence states they face in financial risk workflows, rather than by general capability scores alone.arxiv.org革新
BixBench3 finds AI agents struggle with full-scale computational biology studies
A new benchmark evaluates whether AI agents can turn raw biological data into research outputs across complete computational biology workflows. In 20 tasks, 13 frontier models scored between 0.00 and 0.48, with performance declining as datasets and sequences of analysis steps grew.arxiv.org革新
FLARE uses an LLM and Lean to verify optimization reformulations
Researchers introduce FLARE, a system that pairs an LLM-based agent with the Lean proof assistant to check whether proposed mixed-integer linear programming reformulations preserve the original problem. On the paper’s 20-problem, 109-formulation benchmark, the authors report 100% accuracy on the NP-hard subset and…arxiv.org革新
A preprint proposes profiling AI and workplace tasks by cognitive capabilities
Researchers propose a framework that compares AI systems with workplace tasks using shared cognitive-capability profiles, based on evaluations of six AI systems and task requirements gathered from 410 employees.arxiv.orgセキュリティ
FuzzingBrain-Bench tests whether LLMs can find unexpected software crashes
A new arXiv benchmark evaluates whether large language models can discover distinct crashes in open-source software without being given a predefined vulnerability target. In the authors’ tests, Claude Opus 4.8 triggered crashes in 60 of 77 challenges, but achieved only 196 of a possible 579 points.arxiv.org
毎週 1 回の有益なブリーフィング
フィードに依存せずに AI に追いつきます。
今週の検証済み AI ニュース、オリジナル データ、便利なツール、おすすめの学習情報、最新の AI ジョブを入手します。
AI を学習している人々にリーチする
AI の専門家を雇いますか、それとも便利な AI 製品を立ち上げますか?それを学び、行動するためにここに来た人々の前に置きます。
AI の仕事を投稿するAIツールを提出する