Understanding Off- vs On-Policy Distillation: A Tale of Distinct Training Objectives
Research
Research News
New papers, breakthroughs and academic AI research
Fork-dLLM: Avoiding the Flexibility Trap in Diffusion Language Models
Research
PACE: Policy-Native Adaptive Decision Timing for Long-Horizon Reasoning
Research
Here the World in Stereo: Learning Dynamic Spatial Correspondence for Immersive Joint Video-Audio Generation
Research
Proper Scoring Rule-based Diffusion for Probabilistic Weather Forecasting
Research
Alignment via Training Against Probes Without Losing Monitorability
Research
Provable Test-Time Scaling for Beam Search in LLM Reasoning
Research
Schema: Discovering Unknown Environments via Agentic Program Induction
Research
Markovian Dynamics Enforcer: Feasibility Preserving Correction on Learned Dynamics Manifolds
Research
Alleviating Hallucination in Reasoning Tasks with Training-Free Uncertainty-Guided Steering
Research
Search Shapes Conclusions: Auditing Evidence Selection Bias in Deep Research Agents
Research
Multi-LLM Collaborative Alignment via Stackelberg Games
Research