Local Qwen3.5 GRPO, semantic simplicity and matched length/prompt baselines
Burny
BurnyCoder
AI & ML interests
deep learning, LLMs, interpretability, science, physics
Recent Activity
updated a collection about 2 hours ago
Occam RL: adequate and parsimonious reasoning updated a dataset about 2 hours ago
BurnyCoder/occam-rl-artifacts published a dataset about 2 hours ago
BurnyCoder/occam-rl-artifacts