Benchmark Almanac

University of Chicago

We track 3 benchmarks that University of Chicago made or helped make.

0 in use 3 with public code 1 with a leaderboard here

Benchmarks

Name Kind Released Best score Status
MJ-Bench Models compare image pairs for text alignment, safety, visual quality, and bias. Scores measure preference accuracy, with separate dispersion and equality measures for bias. Image generation and editing Jul 2024 44.5% Not in cards
MMIE Models interpret and generate interleaved text and images. MMIE-Score, a fine-tuned InternVL-2-4B model, grades the resulting multimodal responses. General visual questions Oct 2024 Not in cards
PreScience Models forecast future scientific contributions through collaborator, prior-work, contribution and impact prediction tasks. Real-world events Feb 2026 Not in cards