AI Paper Atlas

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

Non-Parametric Structural Priors for Geometry Theorem Prediction

Junbo Zhao, Ting Zhang, Can Li, Wei He, Jingdong Wang, Hua Huang

ICML 2026 · Large language models › Agents, code and math · approach: Inductive and abstract reasoning evaluation · 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 RAG-retrieved solver-validated candidate theorems to Vanilla ICL (ablation)Pri-TPG (GPT-5 mini) · FormalGeo7K (1400 test)accuracy (%): 26.29 -> 72.64Improved
add TPG precedence prior on top of RAG (ablation)Pri-TPG (GPT-5 mini) · FormalGeo7K (1400 test)accuracy (%): 72.64 -> 84.42Improved
add TPG on hard problems (ablation)Pri-TPG (GPT-5 mini) · FormalGeo7K (1400 test) Hardaccuracy (%): 22.95 -> 40.98Improved
add global prior to Vanilla ICL (ablation)Vanilla ICL · FormalGeo7K (1400 test)accuracy (%): 26.29 -> 29.71Improved
use query-adaptive prior instead of global prior (ablation)Pri-TPG (GPT-5 mini) · FormalGeo7K (1400 test)accuracy (%): 29.71 -> 58.43Improved
use state-aware prior with symbolic validation instead of query-adaptive prior (ablation)Pri-TPG (GPT-5 mini) · FormalGeo7K (1400 test)accuracy (%): 58.43 -> 84.42Improved
add iterative symbolic feedback (vs single-pass) (ablation)Pri-TPG (GPT-5 mini) · FormalGeo7K (1400 test)accuracy (%): 34.3 -> 84.42Improved
increase retrieval pool Top-K from 15 to 200 (ablation)Pri-TPG (GPT-5 mini) · FormalGeo7K (1400 test)accuracy (%): 71.86 -> 84.42Improved
increase Top-K from 30 to 100 (ablation)Pri-TPG (GPT-5 mini) · FormalGeo7K (1400 test)accuracy (%): 79.00 -> 80.29Improved
increase Top-K from 100 to 200 on Hard (ablation)Pri-TPG (GPT-5 mini) · FormalGeo7K (1400 test) Hardaccuracy (%): 30.33 -> 40.98Improved
swap backbone from GPT-5 mini to GPT-5.2 in Pri-TPGPri-TPG · FormalGeo7K (1400 test)accuracy (%): 84.42 -> 89.29Improved
wrap GPT-5.2 in Pri-TPG versus direct solvingGPT-5.2 · FormalGeo7K (1400 test)accuracy (%): 73.14 -> 89.29Improved

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