-
shufanshen/Qwen3-4B-GRPO-DeepMath-50-steps
Reinforcement Learning • 4B • Updated • 16 -
shufanshen/Qwen3-4B-GRPO-DeepMath-100-steps
Reinforcement Learning • 4B • Updated • 14 -
shufanshen/Qwen3-4B-GRPO-DeepMath-150-steps
Reinforcement Learning • 4B • Updated • 19 -
shufanshen/Qwen3-8B-GRPO-DeepMath-50-steps
Reinforcement Learning • 8B • Updated • 17
Shufan Shen
shufanshen
AI & ML interests
Interpretable machine learning, parameter-efficient fine-tuning.
Recent Activity
upvoted a paper about 16 hours ago
On-Policy Parameter Update Direction Underlies Generalization in LLM Post-Training submitted a paper about 16 hours ago
On-Policy Parameter Update Direction Underlies Generalization in LLM Post-Training authored a paper 4 days ago
On-Policy Parameter Update Direction Underlies Generalization in LLM Post-TrainingOrganizations
-
shufanshen/Qwen3-4B-GRPO-DeepMath-50-steps
Reinforcement Learning • 4B • Updated • 16 -
shufanshen/Qwen3-4B-GRPO-DeepMath-100-steps
Reinforcement Learning • 4B • Updated • 14 -
shufanshen/Qwen3-4B-GRPO-DeepMath-150-steps
Reinforcement Learning • 4B • Updated • 19 -
shufanshen/Qwen3-8B-GRPO-DeepMath-50-steps
Reinforcement Learning • 8B • Updated • 17
models 12
shufanshen/Qwen3-4B-GRPO-CodeEurus-150-steps
Reinforcement Learning • 4B • Updated • 20
shufanshen/Qwen3-8B-GRPO-DeepMath-50-steps
Reinforcement Learning • 8B • Updated • 17
shufanshen/Qwen3-8B-GRPO-DeepMath-150-steps
Reinforcement Learning • 8B • Updated • 18
shufanshen/Qwen3-8B-GRPO-DeepMath-100-steps
Reinforcement Learning • 8B • Updated • 13
shufanshen/Qwen3-4B-GRPO-DeepMath-50-steps
Reinforcement Learning • 4B • Updated • 16
shufanshen/Qwen3-4B-GRPO-DeepMath-150-steps
Reinforcement Learning • 4B • Updated • 19
shufanshen/Qwen3-4B-GRPO-DeepMath-100-steps
Reinforcement Learning • 4B • Updated • 14
shufanshen/Qwen3-4B-GRPO-CodeEurus-100-steps
Reinforcement Learning • 4B • Updated • 16
shufanshen/Qwen3-1.7B-GRPO-DeepMath-50-steps
Reinforcement Learning • 2B • Updated • 13
shufanshen/Qwen3-1.7B-GRPO-DeepMath-150-steps
Reinforcement Learning • 2B • Updated • 13