arxiv:2609.05903
Nanxi Li
andyc03
AI & ML interests
None yet
Recent Activity
submitted a paper about 1 month ago
EvoSafeHarness: Evolving Model- and Domain-Specific Harnesses for Securing Agents authored a paper about 1 month ago
EvoSafeHarness: Evolving Model- and Domain-Specific Harnesses for Securing Agents upvoted a paper about 1 month ago
EvoSafeHarness: Evolving Model- and Domain-Specific Harnesses for Securing AgentsOrganizations
None yet
models 19
andyc03/LPG_4B
5B • Updated
andyc03/Qwen3.5-9B-unified-attack-v3
Text Generation • 9B • Updated • 13
andyc03/Qwen3.5-9B-attack-v2.2
Image-Text-to-Text • 9B • Updated • 4
andyc03/Qwen3.5-9B-attack-v2.1
Image-Text-to-Text • 9B • Updated • 7
andyc03/Qwen3.5-9B-v1
9B • Updated • 3
andyc03/Qwen3.5-4B-v2
5B • Updated • 8
andyc03/Qwen3.5-4B-v1.1-allloss
5B • Updated • 4
andyc03/Qwen3.5-4B-v1.1
5B • Updated • 4
andyc03/Qwen3.5-4B-v1
5B • Updated • 5
andyc03/Qwen3.5-4B-attack-lora
Text Generation • Updated • 10
datasets 11
andyc03/attack_data
Preview • Updated • 113
andyc03/latent-policy-guard-40k
Viewer • Updated • 40k • 47 • 3
andyc03/PRISM-DPO-new
Viewer • Updated • 5k • 12
andyc03/PRISM-CoT-new
Viewer • Updated • 10k • 25 • 1
andyc03/VLM-Eval
Viewer • Updated • 1.02k • 32
andyc03/qwen3vl-safedata
Viewer • Updated • 2.43k • 6
andyc03/Eval_Safety-Reasoning-Faithfulness_
Updated • 11
andyc03/robot-training
Updated • 76
andyc03/SDF-Videos
Updated • 64
andyc03/PRISM-CoT
Viewer • Updated • 10k • 69