Keeps 938 of 1,425 evaluated v1 tasks solved at least once by Qwen3-Coder-30B-A3B-Instruct (65.82% pass@4). Archives 487 tasks without observed success alongside 72 prior exclusions. Preserves task payloads and pinned image references. Includes all 5,700 saved conversations, pass-rate and trace-ending plots, reproducible selection, contracts and provenance. One recorded tmux error counts as zero for pass@k while its raw verifier reward remains null. Results come exclusively from the replacement pilot and full run. The README preserves the outstanding v1 Git-history and timestamp leak-check limitation.
Focus BugsInPy v3 READMEs on the new pass@2 evaluation
Preserve all 65 v2 tasks and add 406 newly solved tasks. Include all 16,496 old and new attempts, per-task success counts, success distributions, trajectory statistics, plots, and reproducible task selection.
Align all v1 dataset READMEs with the agreed format
Standardize per-stage explanations, image counts, per-task resource tables and timing coverage across 11 v1 dataset cards. Resource distributions come from the published task configurations. Preserve historical validation gaps and mark unsupported measurements as Missing. No task payloads or archive membership changed.
Document verified original image counts for MiMo and SETA
Restore the compact stats-first README and the standard README, patcher, tasks and archive layout. Embed the image-digest mapping in patch.py. Audit evidence remains accessible at the linked original publication revision. The self-contained patcher reproduces both task and archive Parquets byte-for-byte; 2,509 retained tasks and 1,189 archives are unchanged.
