Fix Day mode theme, add Summary/Paper columns, remove Benchmark Summaries section, enable text selection, format representative samples as readable code blocks

Refactor leaderboard Streamlit app

Productionize project handoff README

Refine space README copy

Merge origin/main into master

Tighten evaluation table and add leaderboard filter

Update leaderboard UX and GRM benchmark content

Support Gradio 5 and 6 runtime config

Reorder leaderboard and add GRM-Bench tab

Refine leaderboard layout and methodology