Epi: Towards a Persistent, Self-Improving Agent for Autonomous Research
Summary
Épi is an autonomous research agent that persists across sessions, learns from accumulated experience, and achieves record-setting mathematical and algorithmic results at reduced compute cost, including 120 records and proofs like Tuza's conjecture for split graphs, plus Épi-turbo for OR coding tasks outscoring Codex on FrontierOR.
View Cached Full Text
Cached at: 10/21/26, 10:21 AM
Paper page - Epi: Towards a Persistent, Self-Improving Agent for Autonomous Research
Source: https://huggingface.co/papers/2610.24944 Authors:
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
Abstract
Épi is an autonomous research agent that maintains persistent investigations across sessions, improves through accumulated experience, and achieves record-setting mathematical and algorithmic results with reduced computational cost.
We present Épi, anautonomous research agentthat sustains investigations across model sessions and evolves its methods through accumulated experience. Épi combinespersistent research history,hierarchical memory, branching and independent review to maintain research continuity, learn from experiments, and explore and revisit alternative approaches. Its experiments run across session boundaries, and afixed verifierevaluates their outputs. Épi produced 120 record-setting results, comprising 20numerical bound recordsacross 12 families and 100 construction records across eight families. These include an approximately 84-fold narrower interval for the n-queens constant and a more than 3.6-million-fold narrower rigorous interval for the quartic Feigenbaum attractor dimension. Exactdual certificatesestablish that theCohn-Elkies linear-programming boundcannot be sharp in dimensions 10 and 11. Beyond numerical improvements, Épi gives acomputer-assisted proofofTuza’s conjecturefor split graphs with an eight-vertex clique part and at most three triangle-active neighborhood types, with arbitrary multiplicities. We extend this research process to algorithmic coding withÉpi-turbo, a lightweight configuration that develops executable solvers for tasks including vehicle routing, scheduling and production planning. On 30FrontierORtasks,Épi-turboachieves a 5.4% higher total benchmark score than Codex at 64% of its model cost, using the same underlying model and reasoning effort. Épi is designed to support human understanding by connecting computational results to explicit methods and mathematical arguments that researchers can examine and build upon.
View arXiv pageView PDFProject pageGitHub9Add to collection
Models citing this paper0
No model linking this paper
Cite arxiv.org/abs/2610.24944 in a model README.md to link it from this page.
Datasets citing this paper0
No dataset linking this paper
Cite arxiv.org/abs/2610.24944 in a dataset README.md to link it from this page.
Spaces citing this paper0
No Space linking this paper
Cite arxiv.org/abs/2610.24944 in a Space README.md to link it from this page.
Collections including this paper0
No Collection including this paper
Add this paper to acollectionto link it from this page.
Similar Articles
EPOCH: Reliable Discovery through Evidence-Governed Search
Researchers introduce EPOCH, an evidence-governed architecture for AI research agents that combines task contracts, typed memory, active falsification, and independent replay, achieving state-of-the-art performance on AlgoTune (0.65 vs 0.53 mean normalized score) and the Math14 suite.
EurekAgent: Agent Environment Engineering is All You Need For Autonomous Scientific Discovery
The paper introduces EurekAgent, an environment-engineered agent system for metric-driven autonomous scientific discovery that achieves state-of-the-art results on math, kernel engineering, and ML tasks with low computational costs.
@THUTeamEureka: 1/3 Excited to open-source EurekAgent! A fully autonomous research system for metric-driven tasks, built with Claude Co…
THU Team Eureka open-sources EurekAgent, an autonomous research system built with Claude Code that achieves state-of-the-art results on math, kernel engineering, and ML tasks through environment engineering.
EviGraph: Evidence-Guided Autonomous Research Agents
EviGraph is an autonomous research framework that represents the research process as a typed evidence graph to maintain claim–evidence consistency across stages, improving claim support rates and experimental data consistency on ARC-Bench-ML and NanoResearch-20.
EvoMaster: A Foundational Agent Framework for Building Evolving Autonomous Scientific Agents at Scale
EvoMaster is a scalable, self-evolving agent framework for large-scale scientific discovery that enables iterative hypothesis refinement and knowledge accumulation across experimental cycles. It achieves state-of-the-art results on four benchmarks including Humanity's Last Exam (41.1%) and MLE-Bench Lite (75.8%), outperforming general-purpose baselines by up to 316%.