Aligned Data Can Induce Misalignment via Context Confusion
Research
Research News
New papers, breakthroughs and academic AI research
Fine-Tuning Diffusion Language Models with Context Selection and Target Weighting
Research
PrefPI: Preference-Guided Steering into Out-of-Distribution Behaviors
Research
Exploiting Vulnerabilities: Universal Adversarial Attacks on Vision-Language-Action Models in Robotics
Research
ArgGYM: A Procedural, Engine-Verified Benchmark for Structured Defeasible Reasoning
Research
HeurEvo: Agentic Evolution of Hybrid Solver-Augmented Heuristics for Time-Critical Mathematical Optimization
Research
DeepRewind: Predicting and Repairing Premature Commitments in Deep Research Agents
Research
Eternal Sunshine of the Spotless Mind: Systematically Erasing LLM's Memories
Research
MemFold: Learning Compact Soft Memory for Long-Context Personalization via On-Policy Optimization
Research
Grounded Revision vs. Prior Injection: Probing Retrieval-Augmented Patent Claim Amendment
Research
Memory Consolidation Flattens the Temporal Shape of User Facts
Research
FinRT: Distilling Adaptive Red-Teaming Strategies into Reusable Adversarial Generators in Consumer Finance
Research