What it measures

Materials-discovery models are compared on structure and energy predictions using Matbench and Matbench Discovery protocols. These are scientific prediction models, rather than a general language-model question-answering test.

Reading the measurement

F1 balances precision and recall. Check both: a system can miss relevant evidence or include unsupported material, and those errors have different consequences. Keep the matching rule, averaging policy and denominator fixed.

Published evaluation

The source leaderboard’s CPS composite is retained on its native scale. This score evaluates interatomic-potential models; it is not an LLM accuracy percentage.

Where this fits in an evaluation suite

A correct multiple-choice answer and a supported scientific contribution demand different evidence. Check whether the score rewards a proposal, an executed experiment, or a completed research target.

Start with a capability hypothesis, then choose a second test with a different failure mode. Keep the checkpoint, reasoning effort, agent tools, sampling policy, and scoring procedure fixed when measuring a training intervention. Record failures and missing outputs in the denominator.

A reproducible comparison

The source leaderboard’s CPS composite is retained on its native scale. This score evaluates interatomic-potential models; it is not an LLM accuracy percentage.

Publish the exact run configuration, item coverage, score units, and available uncertainty. If a source supplies only an aggregate or a single run, preserve that limit. A new model name or a rounded score alone does not establish a capability improvement.

PRIMARY EVIDENCE & FURTHER READING

  1. Matbench Discovery — materials-discovery leaderboard ↗

    Janosh Riebesell

Reviewed 2026-10-26. Explanation and interpretation by RLTasks. Benchmark ownership remains with its original authors.