Datasets and Model Checkpoints for Paper "From SFT to RL: Demystifying the Post-Training Pipeline for LLM-based Vulnerability Detection"
Youpeng Li
Leopo1d
AI & ML interests
None yet
Organizations
None yet
models 6
Leopo1d/OpenVul-Qwen3-4B-GRPO
Text Generation • 196k • Updated • 31
Leopo1d/OpenVul-Qwen3-4B-DPO
Text Generation • 4B • Updated • 2
Leopo1d/OpenVul-Qwen3-4B-ORPO
Text Generation • 4B • Updated • 2
Leopo1d/OpenVul-Qwen3-4B-SFT-ep3
Text Generation • 196k • Updated • 84
Leopo1d/OpenVul-Qwen3-4B-SFT-ep1
Text Generation • 196k • Updated • 2
Leopo1d/OpenVul-Qwen3-4B-SFT-ep5
Text Generation • 196k • Updated • 11
datasets 9
Leopo1d/OpenVul_Sample_Specification_for_RL_Reward_Evaluation
Viewer • Updated • 15.6k • 76
Leopo1d/OpenVul_CWE_Hierarchical_Mapping
Viewer • Updated • 944 • 72 • 1
Leopo1d/OpenVul_Ground_Truth_Vulnerability_Information
Viewer • Updated • 9.77k • 170
Leopo1d/OpenVul_Vulnerability_Query_Dataset_for_RL
Viewer • Updated • 19.5k • 161
Leopo1d/OpenVul_Vulnerability_Preference_Dataset_for_DPO
Viewer • Updated • 7.24k • 45
Leopo1d/OpenVul_Vulnerability_Preference_Dataset_for_ORPO
Viewer • Updated • 7.05k • 34
Leopo1d/OpenVul_Rationalization_based_Vulnerability_Reasoning_Dataset_for_SFT
Viewer • Updated • 15.6k • 13
Leopo1d/OpenVul_Rejection_Sampling_based_Vulnerability_Reasoning_Dataset_for_SFT
Viewer • Updated • 6.28k • 292 • 1
Leopo1d/OpenVul_Distilled_Vulnerability_Reasoning_CoTs_from_DeepSeek-R1-0528
Viewer • Updated • 15.6k • 29