ReasFlow: Assisting Reasoning-Centric Scientific Discovery in Applied Mathematics via a Knowledge-Based Multi-Agent System

2026-07-15

AI Summary

This work addresses the limitations of existing automated scientific discovery systems, which struggle to support theory-driven mathematical research centered on rigorous reasoning and knowledge synthesis due to insufficient scalable logical verification, autonomous exploration capabilities, and procedural heuristic knowledge. To bridge this gap, we propose an end-to-end multi-agent framework that emulates the mentor–student collaboration paradigm, integrating literature review, algorithm design, theorem proving, experimentation, and paper writing within a unified pipeline. Our approach innovatively combines embedded logical verification loops, an automatic knowledge-retrieval-based self-improvement mechanism, and large language model (LLM)-driven orchestration of the research workflow, thereby unifying formal mathematical deduction with procedural heuristics in a single AI architecture for the first time. The system autonomously generates five research papers combining theoretical and empirical contributions, significantly outperforming current open-source baselines under LLM-based peer review, and is publicly released via the ReasLab platform.

Abstract

Recent advances in Large Language Models have fueled autonomous AI agents capable of tackling complex scientific tasks, yet existing automated research systems remain predominantly focused on empirically driven domains with quantitative benchmarks, leaving theory-driven discovery, particularly in mathematically grounded disciplines requiring rigorous proofs and synthesis of domain knowledge, largely underexplored. Key challenges include the difficulty of verifying theoretical reasoning at scale, insufficient reasoning ability for autonomous frontier exploration, and a scarcity of procedural heuristics in the literature. We introduce ReasFlow, an end-to-end autonomous agent system for reasoning-centric scientific discovery that operationalizes a collaborative paradigm where the human expert acts as Principal Investigator while the agent executes rigorous derivations as a capable graduate student. ReasFlow incorporates (i) a robust internal verification loop that audits logical coherence and corrects fundamental errors prior to human inspection, and (ii) an automated knowledge retrieval and self-improvement mechanism that proactively surfaces both declarative facts and overlooked procedural heuristics, substantially reducing expert intervention. The system unifies literature synthesis, algorithm design, theorem proving, experimentation, and manuscript preparation in a single system. Deployed to autonomously generate five complete research papers with rigorous theoretical and empirical content from minimal prompts, ReasFlow consistently achieves the highest evaluation scores among state-of-the-art open-access baselines under a curated LLM-based review rubric. ReasFlow is publicly accessible via the ReasLab platform, providing a collaborative workspace for AI-assisted theoretical research. Github repo: https://github.com/ReasLab/ReasFlow.git.

Problem

• reasoning-centric discovery
• theoretical reasoning
• applied mathematics
• procedural heuristics
• autonomous scientific discovery

Innovation

• reasoning-centric discovery
• knowledge-based multi-agent system
• internal verification loop
• procedural heuristics retrieval
• autonomous theorem proving