arxiv:2505.13291
🔄 In a Training Loop
Michał Wiliński
MWilinski
AI & ML interests
Machine Learning, Reinforcement Learning
Recent Activity
updated a model about 8 hours ago
MWilinski/qwen2.5-3b-dpo-hhrlhf updated a model about 11 hours ago
MWilinski/qwen2.5-3b-direct-hhrlhf updated a model about 11 hours ago
MWilinski/qwen2.5-3b-gailOrganizations
models 32
MWilinski/Llama-3.2-1B-pared-ultrafeedback-deepseek-instruct
Updated
MWilinski/gemma-3-4b-pared-ultrafeedback-deepseek-sft
Updated
MWilinski/qwen2.5-3b-pared-ultrafeedback-deepseek-sft
Updated
MWilinski/gemma-3-4b-pared-ultrafeedback-deepseek-instruct
Updated
MWilinski/qwen2.5-3b-pared-ultrafeedback-deepseek-instruct
Updated
MWilinski/gemma-3-4b-sft-ultrafeedback-deepseek
Updated
MWilinski/qwen2.5-3b-sft-ultrafeedback-deepseek
Updated
MWilinski/gemma-3-4b-direct-hhrlhf
Updated
MWilinski/gemma-3-4b-dpo-hhrlhf
Updated
MWilinski/qwen2.5-3b-dpo-hhrlhf
Updated
datasets 20
MWilinski/ultrafeedback-deepseek-v3.2-irl-scored
Viewer • Updated • 5k • 18
MWilinski/rlhf-irl-pirate-expert
Viewer • Updated • 6k • 11
MWilinski/rlhf-irl
Viewer • Updated • 14k • 72
MWilinski/hh-rlhf-helpful-base-rollouts-gpt-oss-20b-diverse-openrouter
Viewer • Updated • 200 • 63
MWilinski/hh-rlhf-harmless-base-rollouts-gpt-oss-20b-diverse-openrouter
Viewer • Updated • 200 • 93
MWilinski/hh-rlhf-irl
Viewer • Updated • 10k • 6
MWilinski/hh-rlhf-helpful-base-rollouts-gpt-5.1-policy
Viewer • Updated • 2k • 21
MWilinski/hh-rlhf-harmless-base-rollouts-gpt-5.1-policy
Viewer • Updated • 2k • 19
MWilinski/hh-rlhf-helpful-base-rollouts-gpt-5.1-child
Viewer • Updated • 1.5k • 26
MWilinski/hh-rlhf-helpful-base-rollouts-gpt-5.1-adult
Viewer • Updated • 1.5k • 23