Act First, Reason Later: Accelerating On-Policy Distillation for Multi-Turn Agents via Reference-Conditioned Inverse Dynamics
Research
Research News
New papers, breakthroughs and academic AI research
Generalized Residual Closure: General Learning Dynamics for Stability-Plasticity Compatibility
Research
Agent Error Dataset: Scaling 50,000 Error--Diagnosis Pairs for Failure Analysis and Error-Aware Post-Training
Research
Cogentic: Multi-Agent Orchestration for Automated Proof Discovery
Research
SAPO: Single-Rollout Autoregressive Policy Optimization for Agentic Reinforcement Learning
Research
TTMark: Pairwise Distortion-Free Watermarking Beyond Single-Token Entropy
Research
OP-CAD: On-Policy Clean-Audio Distillation for Robust Audio-Visual Reasoning
Research
CHAIN: Calibrated LLM Forecasting via Causal-Temporal Hypergraph Inference
Research
Invariant Atoms: Sparse Coordinates of Local Semantic Geometry in Language Model Representations
Research
Triadic Linear Attention: Three-Dimensional Recurrent States for Long-Context Sequence Modeling
Research
Certified Approximation for Interpretable Representer Landmarks
Research
DAMPER: Return-Prioritized Gradient Control for Smooth Policies
Research