AI Agents Struggle on UniPat AI's SPU Benchmark of Real Wet-Lab Experiment Outcomes
UniPat AI's SPU benchmark tasks AI agents with predicting the outcomes of 4,207 real wet-lab experiments. Across 13 configurations tested, the best average score was 0.486 out of 1, with none exceeding 0.5.
Read the full story at Pandaily ↗
Timeline · 1 report
- 2026-11-15 01:47 · Pandaily
UniPat AI's SPU Benchmark Asks AI Agents to Predict Real Wet-Lab Results, and None Score Above 0.5