How reward hacking emerges in GRPO: Qwen3.5 LoRA adapters for every rollout seed and checkpoint
Rui Wu
RuiWu1123Alignment
AI & ML interests
AI Safety, AI Alignment
Recent Activity
published a dataset about 12 hours ago
RuiWu1123Alignment/rh-dynamics-runs updated a model 6 days ago
RuiWu1123Alignment/Qwen3.5-2B-rh-dynamics updated a model 6 days ago
RuiWu1123Alignment/Qwen3.5-4B-rh-dynamicsOrganizations
