Scoring Higher, Answering Worse: Mitigating Reward Hacking in Rubric-Based RL via Protocol-Level Rubrics
Research
Research News
New papers, breakthroughs and academic AI research
Persistent Watermarking of Text-to-Image Models
Research
Can Vision-Language Models Stay Helpful When Facing Implicit Risks? Intent-Privilege OPSD for Efficient Safety-Helpfulness Alignment
Research
ReSCENE: Server-Side Replay for Structural Mitigation of Catastrophic Forgetting in Federated Continual Learning
Research
Zero-shot Dependency Parsing with Unsupervised Cross-Lingual Bootstrapping
Research
Beyond the Shadows of Plato's Cave: Evaluating False Memory in Autonomous Agents via Counterfactual Reasoning
Research
MotionWeave: Learning Motion-Centered Future Dynamics for Vision-Language-Action Policies
Research
Multi-Link Safety Filtering for VLA Policies Around Moving Hazards
Research
TensorHub: Scalable and Elastic Weight Transfer for LLM RL Training
Research
PARK: Accurate Block Retrieval for Sparse Attention in Video Diffusion Transformers
Research
SparLeak: Privacy Leakage from Sparse Attention in LLM Inference on Shared GPUs
Research
Retrieval Capacity of Self-Attention Under Competition
Research