Luck Is Not Skill: When Do Paired Rollouts Help Group-Relative RL of LLM Agents?
Research
Research News
New papers, breakthroughs and academic AI research
Mind or Message? Auditing Theory of Mind in Multi-Agent Social Simulation
Research
Displacement Geometry Captures Platonic Shared Reality Across Models and Modalities
Research
HyperCLIP++: Fine-tuning CLIP forOpen-vocabulary Semantic Segmentation in Hyperbolic Space
Research
Acceptance-Aware Draft Model Training for Speculative Decoding
Research
Opinion Leader Dynamics: How Sparse Attention Shapes Token Clustering
Research
Per-Query Gating of LLM Rerankers for Multi-Hop Retrieval
Research
Enforcing Narrative Reliability and Epistemic Pacing in LLM-Driven Detective Games via Structured Knowledge Trees
Research
Active Visual Sampling with a Connectome-Constrained Fly Model for One-Shot Hatch Recognition in Architectural Drawings
Research
LLMs as Linguistic Chameleons: Decoupling Semantics and Structure for Privacy-Preserving Communication
Research
PrismGPT: Proxy-Guided Learning for Region-Aware Photo Editing with Self-Synthesized Reasoning
Research
H-Spec: Parallel Speculative Decoding Without a Drafter-Side KV Cache
Research