jacksonlukas/connections-rl-results
connections-rl: raw evaluation artifacts Per-puzzle records, bootstrap summaries and analysis outputs backing connections-rl, a two-scale (Qwen2.5-1.5B / 7B), three-seed study of what verifiable-reward RL actually transfers. This is an artifact bundle for auditing published numbers, not a loadable training dataset, so the dataset viewer is disabled. Read this before using the numbers Two conventions in these files are easy to misread. Both have bitten this project… See the full description on the dataset page: https://huggingface.co/datasets/jacksonlukas/connections-rl-results.
Upload folder using huggingface_hub
Upload folder using huggingface_hub
Upload aug21/data/b1_noscale_ckpt_curve.json with huggingface_hub
Upload aug21/entropy-kl-7b-noscale.png with huggingface_hub
Upload aug21/entropy-kl-7b-noscale.json with huggingface_hub
Upload aug21/entropy_kaggle.ipynb with huggingface_hub
Upload aug21/b3_wandb_export.py with huggingface_hub
Upload aug21/c2_judge_openrouter.py with huggingface_hub
Upload aug21/c2_judge_paired.json with huggingface_hub
Upload aug21/c2_judge_results.json with huggingface_hub
Upload aug21/data/step_count_1p5b.json with huggingface_hub
Upload aug21/data/wandb_train_reward.csv with huggingface_hub
Upload aug21/data/train_stratum_mix.json with huggingface_hub
Upload folder using huggingface_hub
Upload folder using huggingface_hub
Upload aug20/taskD-session-vllm-version.txt with huggingface_hub
Upload folder using huggingface_hub
Update card (generated from committed eval results)
Upload folder using huggingface_hub
Upload folder using huggingface_hub
Upload folder using huggingface_hub
Upload folder using huggingface_hub
Upload folder using huggingface_hub
Upload folder using huggingface_hub
initial commit