- Add styled button bar (GitHub + arXiv) under the title banner
- Link to paper in introduction text
- Direct users to GitHub repo for evaluation data and code
- Add arXiv reference in About tab reproducibility section
Redesign leaderboard with polished HTML tables and styling
- Replace plain Dataframe with styled HTML tables with medals, badges,
highlighted rows, and color-coded deltas
- Add metric summary cards at the top (336/368/566/1270 samples)
- Add gradient header, dark mode support, hover effects
- Remove (Ours) and (Baseline) labels from model names
- Add method filter dropdown for code generation tab
Co-Authored-By: Claude Opus 4 <noreply@anthropic.com>
