Skip to content

Benchmark

ChemBench

Science benchmarks measure how well models answer science questions. Higher scores are better.

Models measured
42
Scores recorded
51
Sources
1
IQ weight
1.45

Score distribution

0.512
Mean
0.516
Median
0.135
Std dev
0.413
10th percentile
0.645
90th percentile
0.0402
Min
0.04020.709
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.45 for chat models, measured over 42 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
51Llama-2-7b-chat-hfLlama-2-7b-chat-hf -base-1741273893· ChemBench verified leaderboard data docs0.0402

Showing 51–51 of 51 results.

Data credits: ChemBench leaderboard, Jablonka group / lamalab-org. Public verified numeric result summaries; results licence not separately stated.