openai/gpt-chat-latest → A (4:3)

Side A changes the ranking color logic from list-position-based gradients to score-based min–max normalization within each group, introducing a dedicated `score_gradient_t` helper, updating rendering to use actual scores, and adding focused tests for normalization, tied scores, and stability. Side B adds several UI enhancements (vote counts, HUD unpin behavior, styling, and tests), but they are more feature-oriented and spread across multiple files, whereas Side A corrects a core visualization behavior with a clearer, reusable design.