# Explore-Execute Chain: Towards an Efficient Structured Reasoning Paradigm

Source: [Explore-Execute Chain: Towards an Efficient Structured Reasoning Paradigm](<https://www.sota2.com/research/paper/68fa7bc71e209131cea8a6a6>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-06-24T05:00:22.883Z

Authors: Kaisen Yang, Tinghe Zhang, Rushi Shah, Kaicheng Yang, Qinwei Ma, Dianbo Liu, Alex Lamb

Published: 2025-09-28T15:48:40.000Z

## Abstract

Many LLMs plan before they act, yet planning and execution are often still entangled in one long generation trace, enforced only through prompts, or split across separate components. We argue that these two stages call for different computation: planning benefits from diversity and breadth, whereas execution demands precision and faithful adherence to a chosen strategy. Treating them as a single undifferentiated chain wastes tokens on routine derivation and makes it costly to explore alternative strategies at test time. We present the \\textbf{Explore-Execute Chain (E\\textsuperscript{2}C)}, which keeps both stages in one model but separates them structurally: a stochastic \\textit{Exploration} phase drafts a concise high-level plan, and a deterministic \\textit{Execution} phase carries it out. Causal SFT and RL train this split so that exploration stays informative and execution remains plan-faithful. Once plans are short yet decisive, extra inference compute can be directed to exploration rather than to repeatedly decoding full solutions. On AIME'2024 at $K{=}32$, \\textbf{E\\textsuperscript{2}C-ReAct Loop} reaches 53.3\\% accuracy with only 12.4k tokens, outperforming Tree-of-Thoughts ($N{=}32$: 50.0\\%, 71.3k). The same structure also supports lightweight domain adaptation: \\textbf{Exploration-Focused SFT (EF-SFT)} updates only the planning phase, uses 3.5\\% of the tokens required by standard SFT, and improves medical benchmark accuracy by up to 14.5\\%.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [Mathematical Reasoning on Algebra (Pass@1)](<https://www.sota2.com/research/sota/mathematical-reasoning-on-algebra-pass-at-1.md>): Mathematical Reasoning; ALGEBRA; Pass@1: 90.9.

- [Mathematical Reasoning on Mathematical Reasoning Aggregate (Pass@1)](<https://www.sota2.com/research/sota/mathematical-reasoning-on-mathematical-reasoning-aggregate-pass-at-1.md>): Mathematical Reasoning; Mathematical Reasoning Aggregate; Average Accuracy (Pass@1): 61.5.

- [Medical Reasoning on MMLU (val)](<https://www.sota2.com/research/sota/medical-reasoning-on-mmlu-val.md>): Medical Reasoning; MMLU (val); Anatomy (AN) Accuracy: 77.

- [Medical Question Answering on MedMCQA](<https://www.sota2.com/research/sota/medical-question-answering-on-medmcqa.md>): Medical Question Answering; MedMCQA; Accuracy: 63.1.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2509.23946>)

[PDF](<https://arxiv.org/pdf/2509.23946.pdf>)
