AI Paper Atlas

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

Deliberate Evolution: Agentic Reasoning for Sample-Efficient Symbolic Regression with LLMs

Xinyu Pang, Zhanke Zhou, Xuan Li, Fangrui Lv, Shanshan Wei, Sen Cui, Bo Han, Changshui Zhang

ICML 2026 · Large language models › Agents, code and math · approach: LLMs for optimization and heuristics · 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 diagnostic tools (data profiler, residual, dimensional) (ablation)Deliberate Evolution (Qwen3-4B) · LSR-Synth PhysicsNMSE: 2.52e-2 -> 4.37e-4Improved
add reflective memory (ablation)Deliberate Evolution (Qwen3-4B) · LSR-Synth PhysicsNMSE: 1.34e-3 -> 4.37e-4Improved
use adaptive operator selection instead of fixed Refine (ablation)Deliberate Evolution (Qwen3-4B) · LSR-Synth PhysicsNMSE: 8.69e-3 -> 4.37e-4Improved
use adaptive operator weights instead of uniform sampling (ablation)Deliberate Evolution (Qwen3-4B) · LSR-Synth PhysicsNMSE: 1.02e-2 -> 4.37e-4Improved
add stagnation monitor (ablation)Deliberate Evolution (Qwen3-4B) · LSR-Synth PhysicsNMSE: 7.69e-4 -> 4.37e-4Improved
add diagnostic tools (ablation)Deliberate Evolution (Qwen3-4B) · LSR-Synth PhysicsAcc0.01 (%): 4.55 -> 15.91Improved
use Deliberate Evolution with 40% sample budget vs LASRDeliberate Evolution (Qwen3-4B) · LSR-TransformNMSE: 1.83e-1 -> 1.15e-1Improved
use Deliberate Evolution vs LASRDeliberate Evolution (Qwen3-4B) · LSR-TransformAcc0.01 (%): 30.91 -> 50.45Improved
use Deliberate Evolution under 1% Gaussian noiseDeliberate Evolution (Qwen3-4B) · LSR-Transform noiseNMSE: 2.46e-1 -> 1.79e-1Improved
use Deliberate Evolution on real-world stress-strainDeliberate Evolution (Qwen3-4B) · Stress-Strain OODNMSE: 6.34e-1 -> 2.98e-1Improved
use Deliberate Evolution on Llama3.1-8B Chemistry accuracyDeliberate Evolution (Llama3.1-8B) · LSR-Synth ChemistryAcc0.01 (%): 16.67 (LLM-SR) -> 11.11Hurt

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