arxiv:2606.00188
Kai Xu
kai-xu
AI & ML interests
Evaluations, safety training, AI for science
Organizations
models 25
kai-xu/OPD-sycophancy-NVIDIA-Nemotron-3-Nano-30B-A3B-BF16
32B • Updated • 3
kai-xu/OPD-sycophancy-gpt-oss-20b
21B • Updated • 5
kai-xu/OPD-sycophancy-Qwen3-8B
8B • Updated • 8
kai-xu/OPD-sycophancy-NVIDIA-Nemotron-3-Nano-30B-A3B-BF16-LoRA
Updated
kai-xu/OPD-sycophancy-gpt-oss-20b-LoRA
Updated
kai-xu/OPD-sycophancy-Qwen3-8B-LoRA
Updated
kai-xu/Qwen2.5-3B-MATH-Translator
3B • Updated • 2
kai-xu/Qwen2.5-3B-MATH
3B • Updated • 6
kai-xu/Llama-3.1-8B-MATH
8B • Updated • 6
kai-xu/Qwen2.5-7B-MATH-Translator
8B • Updated • 8
datasets 9
kai-xu/sycophancy-test-MMLU-multi-turn-v2
Viewer • Updated • 6k • 16
kai-xu/sycophancy-test-MMLU-multi-turn
Viewer • Updated • 6k • 40
kai-xu/sycophancy-train-multi-turn
Viewer • Updated • 14.2k • 23
kai-xu/OPD-sycophancy-IDK-test-MMLU
Viewer • Updated • 2k • 42
kai-xu/OPD-sycophancy-test-MMLU
Viewer • Updated • 2k • 26
kai-xu/OPD-sycophancy-train
Viewer • Updated • 5k • 36
kai-xu/BCT-MMLU-Test-V2
Viewer • Updated • 2k • 20
kai-xu/BCT-MMLU-Test
Viewer • Updated • 2k • 35
kai-xu/BCT-Train
Viewer • Updated • 5k • 13