What it measures

5,520 undergraduate physics problems (in English and Chinese) across 13 physics courses with custom symbolic equivalence verification.

Reading the measurement

Ask whether the evaluation rewards recalling a fact, constructing a model, running an experiment or supporting a discovery. Those tasks require different evidence. For research claims, inspect reproducibility, uncertainty and what was actually checked.

Published evaluation

Published overall physics-problem accuracy for the named language condition. Native metrics are percentages; answer types include numeric, symbolic and other physics responses.

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

Published overall physics-problem accuracy for the named language condition. Native metrics are percentages; answer types include numeric, symbolic and other physics responses.

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. UGPhysics: A Comprehensive Benchmark for Undergraduate Physics Reasoning with Large Language Models — published result table ↗

    Xin Xu, Qiyun Xu, Tong Xiao, Tianhao Chen, Yuchen Yan, Jiaxin Zhang et al. · 2025

    All 9 authors

    Xin Xu, Qiyun Xu, Tong Xiao, Tianhao Chen, Yuchen Yan, Jiaxin Zhang, Shizhe Diao, Can Yang, Yang Wang

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