Reward Under Attack: Analyzing the Robustness and Hackability of Process Reward Models
Paper • 2603.06621 • Published
Udbhav Bamba
udbhavbamba
AI & ML interests
None yet
Organizations
authored a
paper 2 months ago
authored 3
papers 3 months ago
S2D: Selective Spectral Decay for Quantization-Friendly Conditioning of Neural Activations
Paper • 2602.14432 • Published
XRPO: Pushing the limits of GRPO with Targeted Exploration and Exploitation
Paper • 2510.06672 • Published
CRoPS: A Training-Free Hallucination Mitigation Framework for Vision-Language Models
Paper • 2601.00659 • Published • 2