AI Paper Atlas

Research map / Large language models / Agents, code and math

QEDBench: Quantifying the Alignment Gap in Automated Evaluation of University-Level Mathematical Proofs

Santiago Gonzalez, Alireza Amiri Bavandpour, Peter Ye, Edward Zhang, Ruslans Aleksejevs, Todor Antić, Polina Baron, Sujeet Bhalerao, Shubhrajit Bhattacharya, Zachary Burton, John Byrne, Hyungjun Choi and 39 more

ICML 2026 · Large language models › Agents, code and math · approach: Mathematical and structural reasoning benchmarks · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add course-specific rubric penalizing out-of-scope machinery to the judge prompt (ablation)GPT-5.2 Pro judge · QEDBench alignment with human consensusPearson r: 0.69 -> 0.67No clear effect
add course-specific rubric penalizing out-of-scope machinery to the judge prompt (ablation)GPT-5.2 Pro judge · QEDBench alignment with human consensusMAE: 0.13 -> 0.14No clear effect
switch from expert to course-specific rubric for LLM judging of solvers (ablation)Gemini 3.0 Pro solutions · QEDBenchaverage score: 0.957 -> 0.926Hurt
switch from expert to course-specific rubric for LLM judging of solvers (ablation)Claude Sonnet 4.5 solutions · QEDBenchaverage score: 0.883 -> 0.866Hurt
solving constructive combinatorics problems versus procedural ODE problemsClaude Sonnet 4.5 · QEDBench pass ratepass rate: 100.0% (ODEs) -> 27.3% (Combinatorics)Hurt
compare online-available problems with non-online ones (contamination)human-evaluated solver performance · QEDBench 272 problemsmean score / pass rate gap: +0.017 mean-score gap, +0.011 pass-rate gapNo clear effect
control for human proof quality and judge strictness in same-family self-preference regressionLLM judges · QEDBenchself-preference effect: -0.014 to +0.033No clear effect
use strict binary Pass/Fail judge prompts instead of scores (ablation)LLM judges · QEDBenchleniency trendNo clear effect

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy