arxiv:2601.23143
Sangwoo Park
Sangsang
AI & ML interests
I do LLM post-training research (KAIST AI)
Recent Activity
updated a model about 10 hours ago
Sangsang/grpo_Qwen3-0.6B_bs16_g16_mb128_lr1e-6_b1e-3_clip0p2_temp0p7_ep30 published a model about 10 hours ago
Sangsang/grpo_Qwen3-0.6B_bs16_g16_mb128_lr1e-6_b1e-3_clip0p2_temp0p7_ep30 updated a model 1 day ago
Sangsang/feedback_asymmetric_fixed_ema_Llama-3.1-8B-Instruct_bw0p5_fw0p5_ema0p999_ep30_v2Organizations
None yet
models 212
Sangsang/grpo_Qwen3-0.6B_bs16_g16_mb128_lr1e-6_b1e-3_clip0p2_temp0p7_ep30
Text Generation • Updated
Sangsang/feedback_asymmetric_fixed_ema_Llama-3.1-8B-Instruct_bw0p5_fw0p5_ema0p999_ep30_v2
Text Generation • Updated • 13
Sangsang/ci-sft_Qwen3-4B
Text Generation • Updated • 16
Sangsang/ci-sft_Qwen3-4B-Instruct-2507_lr1e-6_ep30
Text Generation • Updated • 14
Sangsang/qwen3_onpolicy
Text Generation • Updated • 32
Sangsang/grpo_Qwen3-1.7B_bs16_g16_mb128_lr1e-6_b1e-3_clip0p2_temp0p7_ep30
Text Generation • Updated • 11
Sangsang/grpo_Qwen3-4B_bs16_g16_mb128_lr1e-6_b1e-3_clip0p2_temp0p7_ep30
Text Generation • Updated • 19
Sangsang/feedback_asymmetric_fixed_ema_DeepSeek-R1-Distill-Qwen-7B_bw0p5_fw0p5_ema0p999_ep30
Text Generation • Updated • 14
Sangsang/feedback_asymmetric_fixed_ema_DeepSeek-R1-Distill-Llama-8B_bw0p5_fw0p5_ema0p999_ep30
Text Generation • Updated • 24
Sangsang/feedback_both_ema_Qwen3-4B_reverse_kl_ema0p999_ep30
Text Generation • Updated • 16
datasets 103
Sangsang/CI-Qwen2.5-32B-Instruct-Augmented-Responses
Viewer • Updated • 729 • 54
Sangsang/CI-Qwen3-32B-Instruct-Augmented-Responses
Viewer • Updated • 729 • 17
Sangsang/CI-Qwen3-32B-Augmented-Responses
Viewer • Updated • 729 • 53
Sangsang/ContextualIntegritySyntheticDataset_Qwen3-32B_all
Viewer • Updated • 729 • 47
Sangsang/safepath_test_2
Viewer • Updated • 39.9k • 27
Sangsang/safepath_test
Viewer • Updated • 39.9k • 27
Sangsang/ContextualIntegritySyntheticDataset_Qwen3-4B_all
Viewer • Updated • 729 • 61
Sangsang/ContextualIntegritySyntheticDataset_Qwen3-8B_all
Viewer • Updated • 729 • 33
Sangsang/CI-Olmo-3-32B-Think-Augmented-Responses
Viewer • Updated • 729 • 35
Sangsang/ContextualIntegritySyntheticDataset_Olmo-3-32B-Think_all
Viewer • Updated • 729 • 34