---
title: "Frontis-MA1: Training an AI4AI Model towards Recursive Self-Improvement in Machine Learning Engineering"
source: "arXiv AI/CL/LG（论文）"
source_tier: "t2"
url: "https://arxiv.org/abs/2607.28568v1"
permalink: "https://feed.astralor.com/items/itm_d480d7554e0a1995f9731b1a"
published: "2026-07-30T17:34:01.000Z"
score: 77
channel: "ai"
tags: ["AI4AI","Frontis-MA1","MLE-Bench Lite","OpenMLE","Recursive Self-Improvement"]
---

# Frontis-MA1: Training an AI4AI Model towards Recursive Self-Improvement in Machine Learning Engineering

## AI commentary

> This paper turns 'self-improvement' from an idea into a runnable engineering system, valuable for both researchers and platform developers. Although the benchmark scale is modest, the open weights and full-stack framework allow the community to build on it. If you care about AI autonomous evolution or agent self-optimization, this is worth reading.

## At a glance

This paper introduces OpenMLE, an open full-stack system for studying recursive self-improvement (RSI) in machine learning engineering (MLE). It includes verifiable environments (OpenMLE-Gym), operator learning (OpenMLE-RL), and long-horizon search (OpenMLE-Evo). The authors post-train Frontis-MA1, a 35B meta-evolution agent, using four atomic program-evolution operators (Draft, Improve, Debug, Crossover) aligned across training and inference. On MLE-Bench Lite under a 12-hour per-task budget on a single RTX 4090 with 12 GB VRAM, Frontis-MA1 raises Medal Average from 39.39% to 60.61% with OpenMLE-Evo, and reaches 71.21% with OpenMLE-Evo-Max, surpassing GPT-5.5+Codex and approaching GPT-5.6 Sol and the 2.8T Kimi K3. On held-out NatureBench Lite, both components transfer, improving Match-SOTA from 50% to 70% with the trained model and from 20% to 50% with OpenMLE-Evo. Model weights and the full stack are open-sourced.

## Key points

- Proposes OpenMLE, an open full-stack system including verifiable environment OpenMLE-Gym, operator learning OpenMLE-RL, and long-horizon search OpenMLE-Evo.
- Frontis-MA1 (35B) trains four atomic program-evolution operators via execution-grounded SFT and RL, coupled with long-horizon search.
- On MLE-Bench Lite, Frontis-MA1 improves Medal Average from 39.39% to 60.61%, and to 71.21% with OpenMLE-Evo-Max, surpassing GPT-5.5+Codex and approaching GPT-5.6 Sol and Kimi K3.
- On NatureBench Lite, both model and framework transfer improve Match-SOTA; weights and code are released.

## Full text

Recursive self-improvement (RSI) requires AI systems that improve the process of building AI (i.e., AI4AI); machine learning engineering (MLE) offers a concrete, executable testbed for studying this capability. We introduce OpenMLE, an open full-stack system for RSI research in MLE, spanning verifiable task environments with execution feedback (OpenMLE-Gym), operator learning (OpenMLE-RL), and long-horizon search (OpenMLE-Evo). On this stack we post-train Frontis-MA1 (35B) as a meta-evolution agent for MLE, aligning post-training and inference around four atomic program-evolution operators (Draft, Improve, Debug, Crossover): the same operators are trained via execution-grounded SFT and RL on data deduplicated against all evaluation benchmarks, then composed into long-horizon search, coupling learning and evolution in a single loop. On MLE-Bench Lite under a 12-hour per-task budget on one RTX 4090 capped at 12 GB VRAM, Frontis-MA1 (35B) improves Medal Average from 39.39% to 60.61% over its base model with OpenMLE-Evo, and reaches 71.21% with OpenMLE-Evo-Max (benchmark-independent experience priors and asynchronous search), exceeding GPT-5.5 + Codex and approaching GPT-5.6 Sol and the 2.8T Kimi K3. On held-out NatureBench Lite, both components transfer: with the framework fixed, swapping in the trained model raises Match-SOTA from 50% to 70%; with the model fixed, swapping in OpenMLE-Evo raises it from 20% to 50%. We release the model weights and the full OpenMLE stack to enable reproducible research on executable AI4AI toward RSI. Code: https://github.com/FrontisAI/OpenRSI

---

Verify against the original before citing; curated by Astral Feed: https://feed.astralor.com/items/itm_d480d7554e0a1995f9731b1a