Archive
Agent、MCP、coding agent、工作流自動化與評測。
最新模型
Building on Qwen3.6-35B-A3B, Thomson Reuters has released Thomson-1.0-Small, which has 35B total parameters and activates 3B per token. The model improves on legal, tax, and research-agent evaluations, but most results were measured by the development team, and the PolyForm Strict license is not a conventional open-source license.
AI 安全
A new study finds that pairing inherently unpredictable questions with authoritative-looking data dashboards makes LLM agents more likely to make directional decisions—even when every figure is fabricated. Fine-tuning on 540 synthetic examples can temporarily repair the action gate, but enforcing a rigid output format causes the improvement to disappear.
AI 安全
Anthropic is bringing its browser agent to all paid plans. Users can approve a plan once, allowing Claude to click, type, and execute workflows across tabs. Every autonomous action is additionally reviewed by a safety classifier, but official testing still found a 0.3% attack success rate, and screenshots may also capture sensitive data.
最新模型
Alibaba has released the weights for Qwen3.8-Flash-Next, which uses Qwen Sparse Attention, Gated Residual, and 51 billion n-gram embedding parameters to reduce the computational cost of long contexts. The model natively supports 262,144 tokens, but its performance figures come primarily from official evaluations, and it is not licensed under the commonly used Apache 2.0 license.
代理工具
The new release can combine multiple LLM scorers using voting, averaging, or custom functions while preserving each constituent score’s rationale and source. The immutable evaluation dataset versions mentioned in the release notes currently support only Databricks; the open-source tracking backend does not yet offer equivalent functionality.
AI 安全
A Beihang University team no longer asks only whether text was generated by AI, but separately identifies the sources of its content development and surface expression. The system outperforms a rerun of RACE on the same data split in the reconstructed MixD2C test set, though cross-dataset generalization and full reproducibility remain limited.
AI 安全
New research finds that guardrails reset after each trajectory cannot detect attack evidence distributed across multiple iterations, while simply decaying risk scores over time can be bypassed by waiting. Across 1,000 attacker-controlled episodes, LoopHarness preserved a 96.9% clean-task completion rate, though the results still require independent reproduction and validation in real-world systems.
代理工具
GitHub is bringing MCP 2026-07-28 to the CLI, SDK, and IDE while adding OpenTelemetry trace correlation across hooks. The new release also removes the legacy `/plugins` entry point, requiring teams to review plugin management, automatic updates, and enterprise policy behavior before upgrading.
AI 安全
New research finds that subagents, persistent goals, and scheduled tasks can repackage tool output as high-privilege user instructions. The attacks achieved all 13 objectives across six agents, and enabling automated permission review did not block the tested attack paths.
最新模型
Built from Common Crawl, LAION-BVD provides open research data spanning video, audio, and frames, including subsets containing 55 million clips with synthetic captions and 300 million frames. ViCLIP outperformed the InternVid baseline by as much as 2.1%, but the data is restricted to research use, while source rights, link rot, and synthetic annotation errors still require separate governance.
AI 基礎設施
AMD has released ROCm 10 and made ROCm.AI—integrating Hyperloom, AMD Skills, and a unified CLI—generally available. The company claims average inference and training performance gains of 3.3× and 2.4× on the same hardware, but those figures cover the entire optimization workflow and should not be interpreted as gains from simply upgrading the runtime.
代理工具
A new flag confines file tools to the working directory and skips user and project settings, making it easier to run agents on evaluation platforms or shared hosts. Tools can still be re-enabled through launch arguments, and it is not an operating-system-level sandbox.