Skip to content

Benchmark

ChemBench Physical Chemistry

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

Models measured
42
Scores recorded
51
Sources
1
IQ weight
1.30

Score distribution

0.595
Mean
0.642
Median
0.214
Std dev
0.327
10th percentile
0.848
90th percentile
0.0242
Min
0.02420.897
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.30 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 docs0.0242

Showing 51–51 of 51 results.

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