arXiv AI
Beyond Plausible Prediction: Safety Sufficiency in Web World Models
Read the original on arXiv AI →The Flow has not summarised this story yet — read it at arXiv AI.
Related stories
arXiv AI
Jun 3
What Benchmarks Don't Measure: The Case for Evaluating Abstention Competence in Autonomous Agents
arXiv:2606. 02965v1 Announce Type: new Abstract: Benchmarks for autonomous agents measure whether agents complete tasks, yet this framing is systematically blind to whether an agent should have proceeded at all.
arXiv AI
Oct 9
Safe Actions Alone Do Not Ensure Safe Agents: Identifying Unfulfilled Obligations with Guard Models
Guard models are used to prevent LLM agents from performing forbidden actions, but this alone does not guarantee safety. The paper shows that many agent trajectories contain unfulfilled safety‑critical obligations—actions that should have been performed—highlighting a new source of risk. To address this, the authors introduce ObligationBench, a benchmark of 240 expert‑validated trajectories, and develop ObligationGuard, a model that improves obligation detection recall and exact‑match rates over existing models.
arXiv AI
Aug 6
SafeCommit: Certifying When Memory-Grounded Agents May Safely Act
arXiv:2608. 04289v1 Announce Type: new Abstract: Long-horizon agents increasingly use persistent memory and tools to take actions with external side effects.
arXiv AI
Aug 18
Toward Safe LLM Agents: A Survey of Specification, Verification, and Enforcement
arXiv:2608. 14590v1 Announce Type: new Abstract: LLM agents increasingly perform irreversible real-world actions, including database updates, API calls, file operations, and autonomous use of tools.
arXiv Machine Learning
Oct 23
EpiSEM-Bench: Evaluating Safety and Evidence Recovery in LLM Agents under Missing Evidence
EpiSEM-Bench is an executable benchmark designed to assess how language‑model agents handle missing safety‑relevant evidence. It comprises 300 matched unsafe/safe task pairs across six domains, systematically suppressing critical facts to measure unsafe execution and failure cardinality. An independent audit tool evaluates whether agents seek missing facts and withhold prohibited actions while completing permissible tasks, revealing variation among ten models in omission robustness, evidence acquisition, and appropriate action.
arXiv AI
6d ago
RePolicy: Learning to Select and Apply Safety Policies for Agent Safeguards
RePolicy is a safeguard that learns to both select the appropriate safety policy from a library and assess agent behavior against it. It processes an agent trajectory and a set of candidate policies, chooses the best-matching policy, applies its requirements, and predicts safety. The authors created PolicyTraj-20K, a dataset of over 20,000 annotated trajectories, and trained RePolicy with supervised and reinforcement learning, achieving top performance on six benchmarks and 19 baselines.