unstable optimization

4 papers

• CodeMerge: Codebook-Guided Model Merging for Robust Test-Time Adaptation in Autonomous Driving
• Enhancing Diffusion-based Unrestricted Adversarial Attacks via Adversary Preferences Alignment
• ReDit: Reward Dithering for Improved LLM Policy Optimization
• Triplets Better Than Pairs: Towards Stable and Effective Self-Play Fine-Tuning for LLMs