AI Paper Atlas

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

ASyMOB: Algebraic Symbolic Mathematical Operations Benchmark

Michael Shalyt, Rotem Elimelech, Ido Kaminer

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 perturbations to seed questionsLLMs on symbolic math · ASyMOBaverage success rate: 74.6% -> 46.8%Hurt
enable code execution (ablation)GPT-4o · ASyMOBsuccess rate: up to 37.7% in a single categoryImproved
enable code execution (ablation)Gemini-2.0 Flash · ASyMOBsuccess rate: up to 8.6% in a single categoryImproved
enable code execution (ablation)frontier models (o4-mini, Gemini 2.5 Flash) · ASyMOBtotal success rate: o4-mini 78.1 -> 76.1; Gemini-2.5 Flash 78.5 -> 69.5No clear effect
enable code execution (ablation)GPT-4.1 · ASyMOBtotal success rate: 48.9 -> 46.2 total; seed 79 -> 83Mixed
replace symbols with numbers (Numeric-All-0, all ones) (ablation)LLMs · ASyMOBsuccess rateHurt
use symbolic instead of one-numeric perturbation (ablation)LLMs · ASyMOBsuccess rateHurt
add equivalence-preserving rewrites of constants (ablation)most LLMs · ASyMOBsuccess rateHurt
use several Easy equivalence perturbations instead of one Hard (ablation)most LLMs · ASyMOBsuccess rateHurt
add the seed question and answer as in-context exemplar (ablation)Gemini 2.0 Flash, GPT-4o, Nemotron Super · ASyMOB Numeric-All-0success rate: +2%, +27%, +43.5%Improved
add the seed question and answer as in-context exemplar (ablation)Gemini 2.0 Flash, GPT-4o, Nemotron Super · ASyMOB Numeric-One-3, Numeric-All-3, Symbolic-3success rate: +2%, +5.1%, +6.8%No clear effect
fine-tune on proofs (DeepSeek-Prover-V2 vs DeepSeek-V3 base)DeepSeek-V3 · ASyMOBtotal success rate: 45.4 -> 46.3No 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