LLMs & Generative AI

A tiny crew of monsters for your Claude Code agents

The Missing Browser for Claude, Cowork & Codex

Turn any website into LLM-ready Markdown

Claude for viral, high-converting social ads

A terminal built for Claude Code w/ one-click session resume

Make existing Codex and Claude Code sessions multiplayer

Wild Pokémon appear while you wait for Claude Code

Any harness, any LLM — cloud-hosted agents on demand.

A Claude Code alternative for people who avoid the terminal

Frontier agent intelligence at Flash prices

Claude Code, Codex & more on your phone

OpenTelemetry for production AI applications

Training large language model agents for long-horizon tool use typically relies on interactions with real or synthesized executable environments, whose construction and verification are costly, or on external simulators that are difficult to ground. We introduce EnvACE, an agentic reinforcement learning method that replaces external environment interaction during training with world rehearsal. The policy alternates between acting and rehearsal: it first generates a tool call, then plays the role of the environment to produce the response induced by that action, and conditions subsequent decisions on the rehearsed response. Both roles are jointly optimized end-to-end using task-success rewards. Through world rehearsal, the policy internalizes the relationship between actions and their environment responses in its parameters, yielding an agent world model that directly supports decision making. Across BFCL-v4, tau^2-Bench, VitaBench, and FinMCP-Bench, EnvACE achieves strong and transferable performance, outperforming environment-scaling baselines in the overall evaluation. Controlled studies further show that world rehearsal consistently improves policy learning across model scales. At test time, the internalized world model enables private rehearsal before committed execution, yielding further gains under a moderate rehearsal budget without additional external interaction. Our findings establish world rehearsal as a new path toward scaling LLM agent training beyond the constraints of external environments. Our code is publicly available at https://github.com/Within-yao/EnvACE.

Economic World Models (EWMs) are generative economic models that simulate how economies evolve from within by modeling heterogeneous agents, their beliefs and actions, and the market and institutional mechanisms through which their interactions produce aggregate outcomes. This paper develops an implementation roadmap for building economic world models as generative engines in which heterogeneous agents act, interact, adapt, and co-evolve with markets and institutions, thereby producing economic dynamics from the inside. We organize EWM systems into a six-level capability ladder, from fixed rule-based agent worlds to adaptive and LLM-based agent worlds, self-evolving agents, evolving institutional worlds, and sim-to-real economic twins aligned with real observations. A systematic literature survey across these levels reveals that existing work remains concentrated in lower-level agent and simulation environments, while systems with self-evolving agents, endogenous institutions, persistent empirical alignment, and validated economic mechanisms remain rare. By translating the EWM agenda into an implementation blueprint, this paper aims to accelerate the development of the next generation of economic simulation environments that can serve as high-fidelity sandboxes for human decision-makers and as training, planning, evaluation, and safety substrates for AI agents. We release a curated paper list and related resources to support future research.

As LLMs are increasingly deployed within agentic systems, their capabilities depend not only on the model weights but also on the harness: the prompts, tools, control flow, memory, and orchestration code surrounding them. This makes automated harness optimization -- the iterative and evaluation-guided improvement of a harness by an AI system -- both an important route to improving AI systems and a demanding capability for AI systems themselves. Yet the community lacks a common protocol for measuring how well frontier LLMs perform at this task. We introduce HarnessOpt-Bench, a benchmark for end-to-end harness optimization under expensive and stochastic evaluation. An optimizer, an LLM paired with a coding harness, receives a target agent's seed harness, graded evaluation feedback, and a fixed target-evaluation budget. It edits the harness and nominates a final candidate, which is scored by its normalized gain over the seed on a held-out test partition that remains inaccessible throughout search. A trusted execution environment enforces the evaluation boundary, meters target-agent resource use, and preserves candidate versions for audit. We evaluate 5 frontier LLMs as optimizers both under a shared coding harness and under their native harnesses across 4 downstream tasks, over 111 scored runs. Experiment results show that optimizer models separate more than the coding harnesses they act through, native harnesses are not consistently superior, and gains vary substantially across tasks and seed regimes. These results establish harness optimization as a measurable and discriminative capability with large space for improvement.

On-Policy Distillation (OPD) is emerging as a promising alternative to reinforcement learning for LLM post-training, yet its effectiveness in multilingual settings remains underexplored. We study OPD and its advanced variant, On-Policy Delta Distillation (OPD^2), for mathematical reasoning in English, Korean, and Japanese. OPD^2 improves OPD by using the probability gap between a post-trained teacher and its base model as the learning signal. Experiments with Qwen3 show that OPD^2 consistently outperforms the original OPD, with particularly strong improvements in Korean and Japanese, and generally narrows the English-Korean performance gap. We further find that English-only OPD can also increase performance for Korean and Japanese, but often shifts the responses toward English, highlighting the importance of multilingual data to preserving target-language responses.

Personalized LLMs with persistent memory are increasingly deployed, yet the faithfulness of their user models remains unexamined. We study over-inference (OI): the phenomenon where LLMs fabricate user attributes beyond what evidence supports. We introduce MirageBench, comprising 150 personas balanced across stereotypical, counter-stereotypical, and neutral profiles, 6 personalization tasks spanning an ``imagination gradient'', a four-way faithfulness taxonomy operationalized by an independent judge (validated against a blind human annotator on 400 claims: Cohen's kappa = 0.863 four-class, kappa = 0.900 binary), and a leaderboard of 12 models across 7 families on 143616 judged claims. We find that over-inference is pervasive: every one of the 12 models over-infers 35%--49% of its claims (cross-model mean 41.6%; claim-weighted 41.8%), with no model in this evaluation escaping it. Most strikingly, we surface a Self-Monitoring Inversion: at the model-selection level, models' self-assessed OI is negatively rank-correlated with their judge-measured OI (rho = -0.60, p = 0.044; exploratory, wide bootstrap CI [-0.90, +0.06], n = 12). The models that report the least over-inference tend to be flagged as fabricating the most, so self-reported confidence is a misleading signal for comparing models, even though within a single model self-audit still ranks that model's own claims moderately well (AUROC 0.58--0.83). We further show that OI is task-dependent (27%--59%) and that, in a multi-turn pilot, inferred attributes accumulate approximately linearly with little revision. MirageBench positions external verification, rather than model self-report, as a more reliable foundation for trustworthy personalization.

LLM agents are increasingly applied to open-ended everyday requests that span work, study, and life. These tasks are long-horizon, cross-environment, and multimodal, forcing the agent to preserve goals and constraints across many steps while navigating heterogeneous tools and attachments. While prior work has addressed individual failure modes such as goals drift, states loss, and context overflow, whether a single harness can manage them jointly and remain effective across backends has received less study. We present OneDayAgent, a long-horizon harness for autonomous agents. OneDayAgent turns an open-ended request into a managed execution process that decomposes tasks into bounded subtasks, maintains execution memory under context pressure, and verifies and repairs the final deliverable. We evaluate OneDayAgent on AgentIF-OneDay across 104 tasks. With the GLM-5.2 backend, OneDayAgent sets a new state of the art with an overall score of 0.821. The same harness runs across five backend LLMs from three model families, indicating the harness generalizes across backends without tuning, even as different models induce distinct execution styles under the same workflow.

Long-horizon reasoning in recent LLMs demands that the model switch between distinct skills inside a reasoning chain, such as first doing a math derivation, then using the result to plan a schedule. We call such problems cross-skill long-horizon tasks: multi-step tasks whose steps require different reasoning skills and depend on earlier outputs. Existing benchmarks often evaluate individual skills, lacking a principled way to measure how well a model switches between skills. We address this gap from both the evaluation and training sides. We introduce Skill Entropy, a measure of the difficulty of switching from one skill to another. We then propose Skill^2-Bench, a benchmark of cross-skill long-horizon tasks built over 558 skills across 9 verifiable and open-ended domains. Each task is assigned a task-level skill-entropy score and grouped into three difficulty levels. Evaluating 8 frontier and 4 open-source models on Skill^2-Bench reveals a skill-switching gap: accuracy decreases on higher-entropy tasks. We then turn skill entropy from a benchmark scale into a training signal. We propose Skill-Entropy RL, an RL framework where the model predicts not only the answer at each step but also the skill used to produce it. The reward combines step-level correctness with a skill-entropy reward that measures the alignment between the model-predicted skill sequence and the gold skill sequence. On Qwen3-4B-Instruct and Qwen3-1.7B, Skill-Entropy RL improves the Skill^2-Bench score from 34.4% to 68.4% and from 14.6% to 40.1%, respectively, outperforming competitive baselines. The same pipeline can be applied to off-the-shelf training data such as OpenR1-Math, indicating that skill entropy is a reusable training signal. Code available at: https://github.com/Gen-Verse/Skill-Entropy-RL

On-Policy distillation (OPD) transfers teacher capabilities by supervising student-sampled trajectories with dense token-level teacher signals. Recent selective OPD methods improve this process by prioritizing signals that are confident, informative, or learnable. However, the assumptions overlook a fundamental failure mode of language models: their token-level judgments can be driven by input-agnostic language priors, formatting conventions, or stereotyped reasoning templates rather than task-specific evidence. We refer to such optimization-relevant but weakly input-grounded supervision as spurious signals in OPD, which may produce large gradients while contributing little task-improving direction. To mitigate this issue, we propose SA-OPD, a Spurious-Signal-Aware On-Policy Distillation framework that identifies and filters misleading token-level supervision based on input-groundedness and optimization impact. SA-OPD introduces a lightweight input-groundedness proxy estimating whether a token-level distillation signal truly depends on the input. It then filters only tokens that simultaneously exhibit low input-groundedness and extreme distillation divergence, thereby removing high-impact spurious updates and achieving fine-grained OPD optimization. Extensive experiments on both large language model (LLM) and vision-language model (VLM) settings demonstrate that SA-OPD consistently outperforms Vanilla OPD and competitive selective methods. These results establish input-groundedness as a key dimension for OPD supervision selection and offer a simple, effective strategy for mitigating spurious updates.

Large language model agents are increasingly evaluated as autonomous tool users, yet most benchmarks focus on bounded tasks with immediate success criteria. Real-world deployments often require Long-Term Coherence, the capacity to preserve purposeful behavior across extended horizons while adapting decisions to accumulated evidence. Evaluating this capacity requires a persistent environment in which actions constrain future choices, feedback arrives at heterogeneous delays, and incoherent behavior produces measurable cumulative effects. Seller-side e-commerce provides a suitable setting for this evaluation through recurrent and interdependent decisions over Product Sourcing, Listing and Pricing Control, Cash-Flow Management, and Mixed-Latency Feedback Adaptation. We introduce MerchantBench, a 365-day order-level simulation grounded in 98,843 real e-commerce product records and equipped with 26 tools for agent interaction. MerchantBench couples promptly observable Upstream Supplier Events with delayed Downstream Order Outcomes, requiring agents to follow individual order lifecycles and revisit earlier decisions. We evaluate eight LLMs under two agent frameworks in 48 runs, each spanning 365 simulated days. Our results reveal a substantial gap between even the latest LLMs and human participants, with the best LLM configuration attaining only 27.3\% of the mean final net assets achieved by human participants.

Language remains an outlier in generative modeling: while images, video, and audio are increasingly modeled in continuous latent spaces, text generation still relies predominantly on discrete tokens. Existing continuous language models either inherit embedding spaces not designed for joint generation and decoding, or compress autoencoded latents to ease diffusion, sacrificing token-level fidelity. Instead of simplifying the representation to suit the generative model, we preserve a high-capacity, decodable text latent and design the diffusion model to learn its distribution directly. We introduce AURORA-LM, a continuous-latent diffusion language model that separates the construction of a decodable text representation from the modeling of its distribution. A Query-based Encoder-Decoder organizes text into a high-capacity, prefix-aligned latent sequence, and a Block-causal Diffusion Transformer learns its distribution through flow matching, generating blocks left to right while denoising positions within each block in parallel. Because such a latent is harder for diffusion to model, AURORA-LM restricts only the noisy-input pathway while retaining the full clean-latent prediction target, accommodating full-width latents without reducing decoder-facing capacity. We further calibrate the noise-level distribution to the latent width, and introduce self-trajectory consistency to bridge independently sampled training noise and iterative denoising at inference. AURORA-LM achieves the strongest performance among evaluated continuous and diffusion-based language models on OpenWebText free generation and XSum summarization. Scaling to 1B parameters with about 1500 EFLOPs of total compute yields further gains, surpassing a larger publicly released latent-diffusion language model under a matched evaluation protocol. All experiments are conducted on Ascend NPUs.

Recent advances in image generation have demonstrated the potential of unified multimodal models that integrate understanding, generation, and editing. However, unified 3D modeling remains constrained by scarce multimodal data, particularly the lack of large-scale and geometrically consistent editing data. To address this limitation, we propose Hunyuan3D-Buffalo 1.0, a unified framework supporting 3D understanding, text-to-3D generation, instruction-guided 3D editing, and text-grounded part generation within a single architecture. To enable scalable training, we construct an 87M-scale 3D multimodal corpus, comprising 25M understanding samples, 50M text-to-3D pairs, and 12M editing pairs generated using Nano3D-v2. Architecturally, the framework combines Hunyuan3D-VLM for semantic, structural, and spatial understanding with Hunyuan3D DiT for high-fidelity 3D synthesis. The VLM provides multimodal semantic conditions for generation, while editing and part generation additionally condition the diffusion process on the source object representation to preserve its overall structure and unedited regions. Extensive experiments show that Hunyuan3D-Buffalo 1.0 achieves state-of-the-art or leading performance on text-to-3D generation and 3D editing benchmarks, while exhibiting strong understanding and part-generation capabilities. Our analysis further shows that both generation and understanding improve editing, demonstrating the effectiveness of unified 3D multimodal training. Project Page: https://tencent-hunyuan.github.io/Hunyuan3D-Buffalo1.0/

We introduce Video-DeepResearch (Video-DR), extending multimodal agents from static images to continuous video streams, a setting that demands dense spatiotemporal grounding coupled with open-web exploration. Preliminary evaluations reveal two critical bottlenecks in current models: (1) modality bias, where agents bypass visual tools in favor of textual search, and (2) parametric knowledge leakage, where models rely on internal memory rather than genuine tool-augmented execution. To address these challenges, we propose Video-DR, featuring a decoupled perception-exploration pipeline with stage-wise tool unlocking that compels exhaustive cross-frame visual grounding prior to web retrieval. Our framework adopts a two-stage training recipe: supervised fine-tuning followed by Group Relative Policy Optimization (GRPO), enabling autonomous exploration that breaks the imitation-learning ceiling. Furthermore, we curate Video-DR-Bench, a human-AI collaborative benchmark comprising 200 complex, multi-hop VQA instances. Empirical results demonstrate that our Video-DeepResearch-35B-A3B establishes a new state-of-the-art of 64.0% average accuracy, surpassing proprietary Claude-4.5-Sonnet (59.0%) by 5.0 points and significantly outperforming GPT-5 (52.5%) and Gemini 2.5 Pro (57.5%). The 30B-A3B variant achieves 59.3%, competitive with Claude-4.5-Sonnet and demonstrating the effectiveness of our training paradigm even at compact scale. Code: https://github.com/Osilly/Vision-DeepResearch.

The Information : Sources: Nvidia agrees to invest $2B in Lancium, the power infrastructure developer of the Stargate campus in Texas, plus $1B more if it hits certain thresholds —  Nvidia has agreed to invest $2 billion into Lancium, the power infrastructure developer behind the OpenAI and Oracle AI campus in Texas …

Arriana McLymore / Reuters : Retailers are updating their websites to rank highly in chatbot results, while making sure purchases are done on their own sites to collect customer data —  As shoppers increasingly turn to ChatGPT and Google's Gemini for product recommendations, retailers are racing to appear in chatbot results …

Axios : OpenAI says it has expanded safety testing around its upcoming model Astra as it “cannot rule out” critical cyber capabilities, potentially delaying its launch —  OpenAI “cannot rule out” that its upcoming model Astra has"critical" cyber capabilities, a designation that has prompted …

The Information : Sources: Canva slashed revenue growth forecast as heavy use of new AI features drove up costs and slowed their rollout, while more Canva users turned to ChatGPT —  Executives at Canva, the design software app with more than 200 million users, had high hopes for a rollout of AI tools that marked …

Mark Gurman / Bloomberg : Sources: OpenAI's new device, slated for 2027, is a hockey puck-sized smart speaker with moving parts that help give it personality, likely costing $300+ —  A highly anticipated new device from OpenAI will have a unique look, complete with moving parts that help give it personality …

Anna Tong / Forbes : Docs: US data labeling companies, like Surge AI and Mercor, that sell training datasets to US AI labs and the government are also selling them to Chinese labs —  The same Silicon Valley startups supplying OpenAI, Anthropic and the U.S. government are also selling AI training datasets …

Bloomberg : Filing: Microsoft recorded $24.1B in revenue from OpenAI during the year ended in June, suggesting OpenAI accounted for more than half of Microsoft's AI sales —  Microsoft Corp. generates most of its artificial intelligence revenue from OpenAI, according to new disclosures from the company.

Jonathan Vanian / CNBC : Meta releases Muse Code in beta, a terminal coding agent powered by Muse Spark 1.2, a coding-focused model priced at $1.25/1M input and $4.25/1M output tokens —  Meta is rolling out its first coding agent called Muse Code as the company tries to challenge leading AI labs Anthropic and OpenAI.

Sam Sabin / Axios : The UK AISI says it observed a total of 19 instances where Mythos and GPT-5.6 Sol tried to hack people and companies during a routine cyber evaluation in July —  The U.K. AI Security Institute said it observed nearly 20 instances of Anthropic and OpenAI's most advanced models trying …

Andrej Karpathy / @karpathy : LLMs are moving from generating artifacts to creating hyper-custom worlds on demand, but still lack the ability to natively perceive and audit what they create —  We're starting to leave the territory where you'd test an LLM by e.g. “create an svg of pelican on a bicycle”.

Lily Hay Newman / Wired : Experts say US law is unprepared for rogue AI agents and models, as recent OpenAI and Anthropic incidents raise questions over legal liability and repercussions —  Both major AI labs' models broke containment, escaped onto the internet, and hacked other companies.

Ben Cohen / Wall Street Journal : A profile of Jacob Tsimerman, who won the Fields Medal last week and is taking a leave from the University of Toronto to join OpenAI and work on AI safety —  Jacob Tsimerman won the biggest prize in math.  Now he's working on the most important problem of his career.

甲骨文今年在数据中心扩建上投入了 700 亿美元，内部也在鼓励员工使用 AI 工具生成代码，但对于 JDK 开源版本 OpenJDK 项目，甲骨文以存在安全、隐私和知识产权风险的理由禁止 AI 生成代码。甲骨文表示，开发者可以私底下使用 LLM 进行代码调试（debug）和审查，但不能将 AI 生成的代码提交到代码库、pull request 或其它项目渠道。

AMD 收购了将权重直接刻在芯片上的 AI 初创公司 Taalas。Taalas 创办于 2023 年，它研发专门用于执行 AI 模型推理任务的集成电路 MSIC（代表 model-specific integrated circuits），其推理速度比 GPU 和其它专用加速器高出一个数量级以上。Taalas 今年 2 月展示了它用台积电 6nm 工艺制造的测试芯片 HC1，测试显示芯片能以每秒 16,960 个 tokens 的速度处理 Meta 的 Llama 3.1 8B 模型，这一速度比英伟达的 GPU 快 48 倍，比 Cerebras 的加速器快 8.5 倍。Taalas 计划今夏推出第二代 HC2 芯片，目标是将参数数量提升至 200 亿。它的缺点是一旦部署只能使用当前的模型，新模型需要重新设计芯片，但并不需要从零开始设计，只需要更换两层金属，因此既便宜也省时间。

字节跳动正在训练一款 10 万亿参数模型。新模型目前正处于预训练阶段。这一过程通常需要耗时三到六个月，随后才能进入微调并最终发布。中国 AI 公司如月之暗面的 Kimi K3 模型有 2.8 万亿参数，而美团的 LongCat-2.0 与 DeepSeek 的 V4-Pro 有 1.6 万亿的参数。Anthropic 最先进的 Mythos 5 参数量估计为 8 万亿，Fable 5 约为 5 万亿。这意味着，单从规模指标来看，字节跳动研发的新模型在规模上已与 Mythos 相近。

《时代》杂志网站开始展示只有 AI 爬虫能看到的带有品牌信息的广告。这些广告旨在影响 AI 聊天机器人，让 AI 从积极的角度谈论品牌产品，而 AI 获得的信息无法追溯到实际页面。《时代》网站排除了 Google 的爬虫，原因是 Google 会惩罚向机器人和真实用户展示不同搜索结果的网站。广告技术平台 Mobian 的联合创始人兼 CEO Jonah Goodhart 指出，影响机器人可能比影响人类用户更重要，因为影响人类用户只能影响一个人。而影响 ChatGPT 就是影响整个 ChatGPT。如果 ChatGPT 改变了它对某个品牌的评价，其影响将是巨大的，远超任何单一广告活动所能达到的效果。

Rust 语言团队宣布了新的 LLM 政策。 除了作者没有人被要求必须阅读 LLM 输出，除非其自愿选择：LLM 输出不允许出现在公开文档、PR 描述或 GitHub 评论中，除非明确标注；审核者如果不愿意，无需查看 LLM 生成的 PR。 没有人必须使用 LLM 为 rust-lang/rust 做贡献：政策必须先为人类编写，仅为机器提供摘要；LLM 审核不能替代人工审核或自审。 你可以生成仅供自己查看的 LLM 内容，无需披露，前提是不会将其发布到任何你期望我们阅读或审核的地方。 你可以用 LLM 回答问题、分析、提炼、完善、检查、提出建议和审核，但不能用来**创造**。

Google 对其 AI 部门 DeepMind 进行了领导层重组，诺贝尔奖得主、DeepMind CEO Demis Hassabis 卸任，他将担任新设立的首席科学家职位并改任董事长。资深工程师以及 Google Brain 联合创始人 Jeff Dean、Sanjay Ghemawat、Oriol Vinyals 和 ​Quoc Le 都离开 Google，他们成立了一家公益性 AI 企业 Discovery Loop，专注于机器学习、科学和工程领域的突破性研究。这次人事变动恰逢 Google 最新 Gemini 模型的发布滞后，原计划 6 月发布，但至今仍未发布，引发了 Google 落后于竞争对手 Anthropic 和 OpenAI 的担忧。

为了推广 AI 工具，许多企业将 AI 使用率作为绩效考核的一部分。结果就是员工为了绩效致力于最大化 AI 使用，导致企业很快发现 token 费用大幅超出预算。软件巨头微软成为最新一家建议工程师限制使用 AI 的公司。微软执行副总裁 Jay Parikh 在一封发给微软员工的邮件中要求工程师专注于业务成果，而非最大化 AI token 的使用量，为了“从 token 投资中获得更大的价值”，微软将比其它模型更便宜的 OpenAI GPT-5.6 设为内部使用的默认模型。自 2026 年 7 月起，微软各部门将设定“AI token 预算目标”，员工可以追踪各自的 AI 支出。

白宫已向美国顶尖 AI 企业透露，在特朗普政府新出台的 AI 安全框架下，中国竞争对手正在开发的开放权重模型将被豁免，无需接受美国政府的安全测试。这项豁免决定是在星期二（4日）的一场白宫闭门会议上向行业代表宣布的。OpenAI、Anthropic PBC 以及 Alphabet 旗下的 Google 等硅谷巨头均派代表出席了此次会议。这一尚未公开的 AI 安全框架，源于美国总统特朗普今年 6 月签署的应对 AI 安全问题的行政令。该行政令提出了一项自愿性计划，鼓励 AI 企业将最前沿的模型交由美国审查。促使华盛顿加速推进安全倡议的导火索，是今年 4 月 Anthropic 警告 Mythos 模型极易发现计算机漏洞，并对该模型的发布实施了严格限制。近几周，OpenAI 和 Anthropic 更接连披露其部分模型曾脱离安全测试环境并入侵第三方机构，进一步加剧了监管的紧迫性。白宫的这项决定对一直呼吁对所有模型实施“强制安全审查”的 Anthropic 首席执行官 Dario Amodei 而言，则是一次重大挫折。

AI 真的会推理吗？它展示的推理链/思维链真的是导出结论前的逻辑步骤？或者它给出的推理过程只是在糊弄我们？一种拟人化的虚构？Quantamagazine 上的一篇文章讨论了这一问题，作者认为 LRM（large reasoning models）在解复杂问题时常常能给出正确答案，但整个机制仍然是不透明的，是一种黑盒子，而思维链更像是某种自言自语而不是严谨的推理。AI 专家 Melanie Mitchell 指出，LRM 有效，执行推理任务的准确率比 LLM 高；思维链不一定忠实反映模型内部的运作过程；模型合成的思维链文本基本无用完全可以删除。2025 年美国东北大学和加州伯克利的研究人员分析了开源前沿 LRM 模型，发现 30%-60% 的思考步骤对模型在数学题上给出的答案“几乎没有因果影响”。

Google 官方博客称，LLM 正在解锁自动化漏洞发现的能力，使得漏洞发现的速度以及数量都远超过去，因此需要采用新的方法应对这一新形势。它的一项应对之策是开发名为动态补丁（dynamic patching）的新功能。该功能可以在大多数情况下让 Chrome 浏览器无需重启就打上安全补丁。利用 Chrome 的多进程架构，动态补丁可以动态地将后台子进程如渲染器和 GPU 的二进制文件逐一替换为更新后的版本。

2026年7月27日，The Peterman Pod主持人瑞安·彼得曼（Ryan Peterman）发布了对朱迪亚·珀尔（Judea Pearl）的长篇访谈。话题从他的科学启蒙、超导存储器与早期人工智能一路延伸到贝叶斯网络、因果阶梯、大语言模型（LLM）和通用人工智能（AGI）。珀尔没有从模型榜单和产品能力出发评价LLM。他把问题放回了一条延续数十年的研究线索：机器究竟怎样把观察变成判断，又怎样从相关性走向行动与解释。

GCC 指导委员会宣布采纳 GCC AI 政策工作组推荐的 AI 贡献政策。GCC 将拒绝任何“包含 LLM 生成的内容或源自 LLM 生成的内容的具有法律意义的贡献”。所谓“具有法律意义”指的是“约 15 行代码和/或文本”即可被视为具有版权意义。GCC 维护者可选择接受由 LLM 生成的具有法律意义的测试用例。AI 政策不禁止将 LLM 用于研究、分析、Bug 发现和报告、补丁审查等用途，只要其输出不包含在贡献中即可。指导委员会表示，AI 政策将不断发展完善，定期进行修订。