Benchmark Almanac

Chinese University of Hong Kong

We track 5 benchmarks that Chinese University of Hong Kong made or helped make.

0 in use 5 with public code 3 with a leaderboard here

Benchmarks

Name Kind Released Best score Status
MemGUI-Bench Agents remember details while working across Android apps. Tasks include recording audio and creating calendar events. A model checks the actions and results across repeated attempts. Phones Feb 2026 73.4% Not in cards
MatCha Models answer multiple-choice questions about material structure, shape and properties using images from materials characterization instruments. Lab and data work Sep 2025 68.1% Not in cards
PRMBench Models identify errors and redundant steps in mathematical solutions. PRMScore combines positive and negative F1 scores across simplicity, soundness, and sensitivity tests. Word problems Jan 2025 68.8 Not in cards
Small Rewriter Benchmark Models rewrite short Verilog designs to reduce hardware use. The authors compare synthesized cell and wire counts and check whether the circuits preserve behavior. Electronics 152.8 Not in cards
MMDM Decision models assign probabilities to options in 5,448 questions across 16 categories. Hard choices use accuracy or exact set match, while soft targets use Brier error. Language reasoning Nov 2026 Not in cards