Archive
Agent、MCP、coding agent、工作流自動化與評測。
代理工具
Hugging Face has added native OpenEnv integration to TRL, allowing GRPOTrainer to access multi-turn environments through a unified factory interface instead of treating each tool call as an independent function. Environments can run as WebSocket services or inside containers, but the project remains experimental, and users must verify isolation strength and reward integrity themselves.
AI 基礎設施／分散式運算
Google has unveiled a complete AI library path for running Ray on TPUs, using a topology field to ensure model workers are placed within the same ICI slice. The new release also adds native JAX data batching, JaxTrainer, official containers, and TPU monitoring, though GKE remains the primary validated environment.
代理工具
Researchers from Microsoft Research and other institutions have introduced OpenForgeRL, enabling Codex, OpenClaw, and GUI agents to undergo end-to-end reinforcement learning within their native execution frameworks. The system uses a model-call proxy server and Kubernetes to isolate rollouts, but its code, data, and weights have yet to be fully released.
最新模型
A KAIST team has proposed ReMo, a training-free method that removes visual tokens at the input stage when their content can be explained by audio, while preserving object semantics through text proxies. In the authors’ own tests, it compressed inputs to the Qwen Omni series by more than half without reducing average scores, but neither the code nor independent reproduction results have been released.
具身 AI／機器人
Black Forest Labs and mimic robotics connected FLUX 3’s video world model to a lightweight action decoder, using the same representations for visual prediction and robot control. The team reported a 95% success rate in an internal deformable-parts sorting test, but the model, data, and complete evaluation have not yet been released.
最新模型
Hugging Face and BigCode have released a new generation of open code pretraining data containing approximately 173 million training samples and 4.9 trillion tokens. The new version no longer requires users to retrieve the source code separately, but training teams must still manage licensing, data contamination, and the 15.9 TB storage cost themselves.
最新模型
Kwaipilot has released the KAT-Coder-V2.5-Dev weights, fine-tuning Qwen3.6-35B-A3B on 127,000 samples before applying reinforcement learning on agentic tasks. The team reports 69.40% on SWE-bench Verified in its own evaluation, although the score is highly sensitive to the agent framework and tool configuration.
代理工具
NVIDIA has updated NemoClaw’s installation process for DGX Station and DGX Spark to prevent existing vLLM workloads from being stopped accidentally and to improve credential and rate-limit handling for large model downloads. The release remains in alpha, and the newly supported DGX OS configurations have not yet completed end-to-end qualification.
開發者平台
DeepSeek retired the `deepseek-chat` and `deepseek-reasoner` compatibility names on July 24, meaning API requests that have not been migrated may now fail outright. The new interface separates model selection from reasoning mode, making the upgrade more involved than simply replacing a model string.
GitHub Repo
Andrew Ng’s team has released OpenWorker under the MIT License, keeping the agent loop, credentials, and conversations on the user’s computer while supporting cloud models and local models via Ollama. It can generate files and operate Slack, calendars, and terminals, but “local-first” does not necessarily mean data never leaves the device.
最新模型
InclusionAI has released Ling-3.0-flash, a 124B-parameter model that uses KDA, MLA, and a highly sparse MoE architecture for long-horizon agentic reasoning. The API is now available, but the weights are not scheduled for release until August, so its performance claims cannot yet be fully reproduced.
最新模型
Google is moving Gemini 3.6 Flash and 3.5 Flash-Lite to stable APIs, targeting complex primary agent workloads and high-throughput subagents, respectively. Official data shows improvements in token usage, coding tasks, and computer use, but most comparisons come from Google or its evaluation partners and still require validation in real-world workflows.