Papers arxiv:2605.14454 LiSA: Lifelong Safety Adaptation via Conservative Policy Induction Published on May 14 Submitted by minbeomkim on May 15 Google Authors: Minbeom Kim , , , , , , , , Abstract LiSA enables adaptive safety guardrails for AI agents by converting occasional failures into reusable… …
… We further introduce ICU -Evo to study structured-memory agents that improves long-horizon reasoning but does not fully eliminate safety failures. Together, Real ICU provides a clinically grounded testbed for measuring and improving AI sequential decision-support in high-stakes care. …
… The following papers were recommended by the Semantic Scholar API SaFeR-Steer: Evolving Multi-Turn MLLMs via Synthetic Bootstrapping and Feedback Dynamics 2026 ContextualJailbreak: Evolutionary Red-Teaming via Simulated Conversational Priming 2026 Transient Turn Injection: Exposing Stateless Multi-… …
… When the models are instructed to prioritize specific values that conflict with their own preferences, they often fail to override their default actions, choosing incorrect actions for 80% of the time. …
… We formalize this perspective as World-Action Interactive Models WAIMs , and instantiate it in autonomous driving with DAWN Denoising Actions and World iNteractive model , a simple yet strong latent generative baseline. …
… Concern-Level Diagnostics for AI Peer Review 2026 Evaluating Patient Safety Risks in Generative AI: Development and Validation of a FMECA Framework for Generated Clinical Content 2026 PhysicianBench: Evaluating LLM Agents in Real-World EHR Environments 2026 An Empirical Study of Agent Skills for He… …
… The following papers were recommended by the Semantic Scholar API PISmith: Reinforcement Learning-based Red Teaming for Prompt Injection Defenses 2026 Uncovering Linguistic Fragility in Vision-Language-Action Models via Diversity-Aware Red Teaming 2026 T-MAP: Red-Teaming LLM Agents with Trajectory-… …
… Leveraging the observation that driving VLAs often emit structured JSON-like outputs, Fast-dDrive freezes structural tokens into a section scaffold and employs a section-aware training recipe that prioritizes safety-critical planning. …
… We study a 21-day deployment of 3,505 user-funded agents trading real ETH onchain. The system logged 7.5M agent invocations, ~300K onchain actions, ~$20M in volume, 70B inference tokens, and 99.9% settlement success for policy-valid transactions. …
… PR: github.com/google-gemma/cookbook/pull/342 HDP spec: helixar.ai/about/labs/hdp HDP-P spec: helixar.ai/about/labs/hdp-physical · cc @ sergiopaniego thats amazing This exactly fits our kid education domain! …