Bring results into the first screen with the Challenge website’s light/dark styling, a five-metric table and the shared mountain visualization. Add an all-results view while preserving paired-only comparisons, DSH provenance and expandable scoring evidence.
Validated all 13 runs, five sorting metrics, 6 baseline / 7 harness / 6 paired views, search, Chinese/English, light/dark themes and narrow layouts. Published data, scoring and result archives are unchanged.
Add DeepSeek API + DSH results and independent leaderboard entry (#6)
- Archive DeepSeek API + DSH full-1000 results and recovery provenance (6ea7c5684664bfe5c6e3d282a5e28f68244acd9e)
- Display DSH as an independent harness result with verified metrics (405b3c1060498dcd94a09890ff7a48162c9cedf2)
Co-authored-by: RenyuanLi <lry520111@users.noreply.huggingface.co>
Refresh leaderboard with single-pass Agent and DeepAgents comparisons (#4)
