arxiv:2606.04923
Xuekang Wang
wxk123
·
AI & ML interests
LLM Safety
Recent Activity
new activity 3 days ago
Alibaba-AAIG/SEAGLE:关于 SEAGLE 与 Speculative Safety-Aware Decoding (EMNLP 2025) 的技术关联讨论 new activity 3 days ago
Alibaba-AAIG/SEAGLE:关于 SEAGLE 与 Speculative Safety-Aware Decoding (EMNLP 2025) 的技术关联讨论 authored a paper 3 months ago
Reproducing, Analyzing, and Detecting Reward Hacking in Rubric-Based Reinforcement LearningOrganizations
models 6
wxk123/llama-3.2-1b-instruct-augmented
1B • Updated • 425
wxk123/TinyLlama_augmented
1B • Updated • 8
wxk123/llama2-7b-chat-aug_FTattack_SFT_noWarmup_5epoch
Updated
wxk123/llama2-7b-chat-aug-underFTAttack_SFT_noWarmup
7B • Updated • 3
wxk123/llama2-7b-chat-aug-underFTAttack_softSFT_withWarmup
7B • Updated • 3
wxk123/llama2-7b-chat-aug
7B • Updated • 3 • 1
datasets 0
None public yet