Scan record
Nov 1839 loaded
- Letting Disagreement Surface Beats Resolving It: A Gated Ablation Study of Provenance-Weighted Memory in Multi-Agent Debate
PaperarXiv: 其他 AI 安全论文The item you came from
- Visual backdoors grow as student model rollouts continue
PaperarXiv: 其他 AI 安全论文
- Bad Apples in the Kitchen: How LLM Teams Detect, Correct, and Calibrate Against Misaligned Members
PaperarXiv: 其他 AI 安全论文
- Cloaked in authority: a practical black-box jailbreak of MLLMs exploiting visual trust bias
PaperarXiv: 其他 AI 安全论文
- ConfGuard: A Physics-Grounded Reference Monitor for Predictive Maintenance LLM Agents
PaperarXiv: 其他 AI 安全论文
- DACE: Diversity-Driven Adversarial Co-Evolution for Robust LLM Safety Alignment
PaperarXiv: 其他 AI 安全论文
- From token changes to task failure: bit flips in shared KV cache blocks
PaperarXiv: 其他 AI 安全论文
- GapLeak: when privacy-protecting LLM web agents become a privacy risk
PaperarXiv: 其他 AI 安全论文
- EGS: a supplementary safety framework for educational LLMs
PaperarXiv: 其他 AI 安全论文
- Is 4-bit Quantization More Than Noise for Chat Model Safety Behavior? Attributing Behavior Change with Matched Perturbations
PaperarXiv: 其他 AI 安全论文
- Can You Undo That? Reversibility Grading of AI Agent Operations
PaperarXiv: 其他 AI 安全论文
- MedGuardBench: A Benchmark for Developing Patient-Facing Personalized Guardrails
PaperarXiv: 其他 AI 安全论文
- WebAgents: splitting agent creation, hosting, and execution for the agent-native Web
PaperarXiv: 其他 AI 安全论文
- Wrong Inputs, Wrong Thresholds: Re-evaluating Adversarial Face Privacy
PaperarXiv: 其他 AI 安全论文
- Auditors struggle to detect deliberate sabotage from end-to-end records of AI safety research
PaperarXiv: 其他 AI 安全论文
- The Threat Landscape of Vibe Coding and the Path to Safe Autonomy
PaperarXiv: 其他 AI 安全论文
- Chain-of-thought controllability reduces chain-of-thought monitorability
PaperarXiv: 其他 AI 安全论文
- Disentangling Moral Circuits and Moral Features in Language Models: Moral Content, Normative Judgment, or Cue Word? Evidence from 14 Models
PaperarXiv: 其他 AI 安全论文
- Depth-Aware Activation Steering Based on Attention Residuals
PaperarXiv: 其他 AI 安全论文
- Evaluating semantic insensitivity in financial sentiment analysis: implications for sustainable financial decision-making
PaperarXiv: 其他 AI 安全论文
- From preference data to personalization: tracing sycophancy in LLMs
PaperarXiv: 其他 AI 安全论文
- The defense rests: a court-inspired multi-agent LLM framework for automated safety argument generation
PaperarXiv: 其他 AI 安全论文
- "Who will take over, and where the company goes from here": analysts expect the uncertainty facing investors to persist after Nvidia's earnings beat
NewsCNBC · TechnologyAI updates
- Fireworks updates its investigation of two October 2026 security incidents; a small number of customer keys, models and datasets were accessed
NewsFireworksEvent
- DarkRange says n8n MCP OAuth has a zero-click open redirect vulnerability
NewsDarkRangeEvent
- NewCore says two Grok Bots made 31 uninstructed requests for enterprise data over five days, all blocked
NewsNewCoreEvent
- OpenAI launches dual GPT-6.2 models, Astra first to reach the high-risk threshold for AI self-improvement
NewsiThome (Taiwan)Evaluation
- After Jensen Huang's departure, interim CEO Cress shows NVIDIA remains at the center of AI
NewsCNBC · TechnologyAI updates
- Microsoft and more than 40 organizations launch the AI Transparency Initiative to develop open standards
NewsSource EMEAGovernance
- U.S. Secretary Malin said the recent system failures were not cyberattacks but stemmed from aging facilities and upgrade pressure
News澎湃新闻 · 检索「AI风险」AI updates
- Jensen Huang is alleged to have had an extramarital affair with a Chinese spy
NewsPlatformerAI updates
- Brazil's National Data Protection Authority (ANPD) AI regulatory sandbox receives an international award in Brussels
NewsAgência Nacional de Proteção de DadosGovernance
- Coinbase says its AI intercepted $4.7 million in transfers that would have gone to a deepfake scam
NewsDiarioBitcoinEvent
- Former OpenAI board member Helen Toner warns of AI self-improvement risk, saying the alignment challenge remains far from solved
NewsTVNOpinion
- South Korea tightens exam security after AI cheating scandal
NewsK-SélectionEvent
- Study shows some AI models change their answers when told they are being monitored
NewsScience et vieAlignment
- California attorney general sues OpenAI over agent breach of Hugging Face
NewsGIGAZINEGovernance
- Keynote speech by Singapore Senior Minister of State Tan Kiat How at the StableX Agentic Network launch event
NewsMDDI (Singapore)Governance
- Gemini conversation retention period to be extended from 72 hours to 2 months
NewsiThome (Taiwan)AI updates