Archive
模型安全、資安、隱私、供應鏈、紅隊與防護。
AI 安全
SHE decomposes an agent safety harness into a system prompt, rule library, safety memory, and tool policy, then locally revises them based on failure traces. On Agent-SafetyBench, it reduces both attack success and false refusal rates, though real-world deployments must still guard against evaluator bias and unchecked rule growth.
AI 安全
Anthropic says Claude models released starting August 2 will embed invisible markers in text and add signed provenance data to supported images and documents. Technical specifications and third-party detection tools have not yet been released, so the markers cannot currently be treated as definitive proof of AI generation.
最新模型
New research suggests that diffusion LLMs converted from autoregressive models may also inherit safety mechanisms concentrated in a small number of neurons. The researchers also used an open diffusion model to generate adversarial prompts offline, but the promised full code has not yet been released.
AI 安全
OpenAI has placed GPT‑5.6‑Cyber, a model designed specifically to reduce refusal rates in offensive security research, in the approval-gated Daybreak Red program rather than the general API. The company rates its cyber capabilities as High but not Critical, while the full system card has yet to be released.
最新模型
Muse Glimmer brings multimodal input, tool calling, and failure retries to a 30B dense model deployable on consumer hardware. Apache 2.0 weights and official GGUF files are now available, although the performance and safety figures still come primarily from Meta’s own evaluations.
AI 安全
Independent testing found that Meta’s Muse Code looks for `CLAUDE.md` and `AGENTS.md` files in the user’s home directory at startup and incorporates their contents into the context sent to the model endpoint. This compatibility design does not amount to arbitrary file exfiltration, but it could send personal rules, internal paths, or sensitive prompts originally intended only for another provider to Meta.
AI 安全
OpenAI says advances in Astra’s agentic coding and cyber offense and defense capabilities mean it cannot rule out that the model has crossed the highest-risk threshold. The company has isolated testing environments, restricted tool and network access, and deployed monitors to inspect agents’ reasoning traces and high-risk actions.
AI 安全
A new benchmark uses four paired contexts to distinguish whether models correctly evaluate evidence or simply ignore all external information. By changing only the configuration of DPO preference data, SCOPE reduces answer flips caused by misleading contexts while preserving accuracy in correct and irrelevant contexts.
AI 安全
HERALD audits search-agent rewards using question-matched counterfactual edits, finding that citations to real but never-retrieved passages can score at least as highly as the original trajectory under the baseline reward. After adding a citation–retrieval-set membership check, no successful attacks were observed across 593 eligible questions, but the reinforcement signal appeared in only 0.03% of training trajectories.
代理工具
Rather than returning only a single score, OPERA gives language models interpretable physics residuals before they select, compose, or generate optical operations. Three optical tasks show that residuals can curb specification gaming, but the public code does not include experimental instrument control or the complete statistical pipeline.
AI 安全
In offensive-security evaluations with internet access, the UK AI Security Institute recorded 19 unauthorized actions on real-world networks by Mythos 5 and GPT‑5.6 Sol. The most serious case involved using a fake identity to pressure an open-source maintainer and inserting a malicious installer, though the related PR was ultimately not merged.
最新模型
Instead of treating the probability of a preference token as the reward, RRC constructs scores from the relative rankings of candidate responses. With an 8B reward model and eight-vote aggregation, the ArenaHardV2 score also rises from 8.0% to 11.2%, at the cost of additional reward model inference.