Preprint 2026 · Open Full-Stack Release · OpenRSI First Release

Frontis-MA1: Training an AI4AI Model towards
Recursive Self-Improvement in ML Engineering

Horizon Research, Frontis.AI1  ·  Tsinghua University2

Open weights · open gym · open search — the full OpenMLE stack, released

📄 Citation </> Code 🤗 Models 🧩 Tasks Results ↓
LIVE  epoch 1 nodes 0 best score 0.000 operator draft
Draft Improve Debug Crossover failed run ◉ incumbent best

Live · illustrative simulation, not paper trace. OpenMLE-Evo growing a population of executable programs — node size ∝ simulated validation score. Move your cursor to perturb it.

Project Film · 64 seconds

See the complete OpenMLE stack
in one minute.

From executable tasks and execution-grounded learning to evolutionary search and reported results: a visual tour of OpenMLE-Gym, OpenMLE-ERL, OpenMLE-Evo, and Frontis-MA1.

H.264 · 3840 × 2160 · no audio Open MP4 ↗

01 · TL;DR

AI that improves the process
of building AI

One loop

Draft · Improve · Debug · Crossover — trained with execution-grounded SFT + RL, then composed into long-horizon search. Learning and evolution close.

39.39 → 71.21

Base → post-trained → Evo-Max, MLE-Bench Lite @ 12 h / task on one RTX 4090 capped at 12 GB VRAM. Beats GPT-5.5 + Codex (68.18), approaching GPT-5.6 Sol and the 2.8T Kimi K3.

100% open

Weights · gym · sandbox · training · search · eval. The full stack, released.

Full abstract

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.

0%
Medal Average — above GPT-5.5 + Codex (68.18%)
+0pp
Post-training gain over base, identical harness
0
Quality-gated executable tasks in OpenMLE-Gym
0
Execution-verified SFT samples
0
Trainable atomic operators = the action space
0B
Parameters — Pareto frontier vs trillion-scale systems

02 · Mechanism Ladder

When the improver is trained,
evolution becomes meta-evolution

Verified trajectories return to train the model that generates future evolution.

L1

Evolution

Artifacts evolve. The improver stays frozen.

L2

Self-Evolution

Experience flows back into the search.

L3

Meta-EvolutionTHIS WORK

The improver itself is trained — SFT + RL on verified trajectories.

L4

Recursive Self-Improvement

Each system improves the process that builds its successor.

Frontis-MA1 the operator model · the improver TRAINING SIDE OpenMLE-Evo experience-guided long-horizon search INFERENCE SIDE verified trajectories → SFT + RL operator weights → drive search META-EVOLVING

03 · The Open Full-Stack

Unit of training = unit of search.
One stack closes the loop.

Sandbox feedback = training signal = search signal.

MODULE 01 — ENVIRONMENT

OpenMLE-Gyma gym, not a dataset

≈11,000 · Meta Kaggle catalog 3,972 · eligible · 36% 2,839 · executable · 26% 2,240 · quality-gated · 20% ✓
  • 5,758 executable tasks · 3 sources · MLE-Bench excluded
  • 6 structured sandbox feedback modes

MODULE 02 — TRAINING

OpenMLE-ERLexecution-grounded SFT + RL

W ×4.0 rollout group · bound W clips · entropic top-weighting
  • 26,259 execution-verified SFT samples
  • Async rollouts · no straggler stalls

MODULE 03 — INFERENCE

OpenMLE-Evotest-time scaling → test-time learning

U = λs·s̃ + λΔ·Δ̃ + λn·ν s̃① Δ̃① ν① score leader gain leader novel family Improve P(gain leader): 10.47% → 17.09%
  • 1 experience card / node → task-global board
  • On-demand, operator-conditioned memory

Frontis-MA1-30B / 35B

Trained by OpenMLE. Driving OpenMLE. Evaluated only on third-party benchmarks.

PRODUCT × ENGINE

04 · Trainable Atomic Operators

Four operators. One unified
action space for code evolution.

Invoked thousands of times per task — repair, refine, recombine.

Draft

generate from scratch

Improve

refine a parent program

Debug

repair failing code

Crossover

recombine two parents

① Evolutionary Inference — operator-driven expansion of the solution tree

Root · Task & Data 0.61 0.47 Error 0.58 Error 0.32 Error 0.67 0.73
Draft Improve Debug Crossover edge style = operator · node value = sandbox validation score
Inference Loop 1Choose Operator 2Select Parent 3Generate Code 4Sandbox Execute 5Evaluate & Rank 6Update Database ⟲ iterate

Scaffold-agnostic → Greedy · abMCTS · AIRA-EVO · OpenEvolve


05 · MLE-Bench Lite · 12 h / task · 1× RTX 4090 · 12 GB VRAM cap · avg of 3 runs

A 35B open model at the table
of frontier systems

Qwen3.6-35B-A3B base · OpenMLE-Evo
39.39
Frontis-MA1-35B post-trained · OpenMLE-Evo
60.61
Claude Opus 4.8 Claude Code
63.64
GPT-5.5 Codex
68.18
Frontis-MA1-35B OpenMLE-Evo-Max · priors + async parallel search
71.21
GPT-5.6 Sol / Kimi K3 Codex / Claude Code · up to 2.8T
72.73

Gains compose everywhere: 30B 34.85 → 53.03 → 66.67 (+18.18pp) · GLM-5.2 59.09 → 66.67 · MiniMax M3 54.55 → 65.15 · Kimi K2.6 59.09 → 66.67 · medal mix shifts toward Gold — matching Kimi K3's Gold rate.

AIRA-Evo baseline · tokens / 12h run

129.3M

Free-form history — every expansion gets more expensive.

OpenMLE-Evo · same checkpoint, seed & budget

75.3M −41.7%

Prompt −50.3% · nodes −12.4% · more new-best updates.

BRONZE — leaf-classification

Late Crossover + Improve jumps

0.9455held-out Human Rank

Debug → Crossover → ConvNeXt leap. 85.0% of the gain from late ops. Best baseline: 0.6303, no medal.

SILVER — mlsp-2013-birds

Memory-guided recombination

0.8889held-out Human Rank

Memory keeps winning branches, flags a weak ResNet50. 91.9% of the gain from Improve + Crossover. Best baseline: 0.2963, no medal.


06 · RL Training Curves · Frontis-MA1-35B

Training signal rises through
execution-grounded RL

The rollout reward is noisy by design, while held-out validation and medal count expose the underlying learning trajectory. Hover the chart for exact checkpoint values.

221

RL policy updates logged

0.180 → 0.400

Rollout reward · first to final step

0.404

Peak validation base reward

31 /176

Peak validation medal count · out of 176 tasks

Reward / medals
stitched training step

Left axis · base reward Right axis · medal count Source · qwen36_rl_stitched.csv · inlined

07 · Meta-Ability & Transfer

It learned how to improve
not one leaderboard's tricks

NatureBench Lite, controlled: hold the model or the framework fixed — both transfer.

Model transfer · framework fixed · Match-SOTA

50%70%

Same framework, Frontis-MA1 swapped in.

Framework transfer · model fixed · Match-SOTA

20%50%

Same model, OpenMLE-Evo search swapped in.

5 / 5

Every modality group improves. No Medal Rate drops.

+2 +4 +1 +4 +3

14 new medals, spread across all five modalities.

24.2 → 34.8

Test medal rate with upper-tail reward shaping (×4.0).


08 · Open-Source Release

Models, tasks, training, and search.
A reproducible OpenMLE stack.

R-01

Frontis-MA1 weights

30B & 35B post-trained checkpoints on HuggingFace.

R-02

OpenMLE Tasks

Audited task inventory + category-specific release artifacts.

R-03

Sandbox infrastructure

Scheduler + CPU/GPU Docker workers; control / execution / storage separated.

R-04

OpenMLE-ERL

Execution-grounded SFT data construction + asynchronous RL configs.

R-05

OpenMLE-Evo

Experience cards, 3-factor selection, operator-conditioned context.

R-06

Evaluation harness

Full MLE-Bench Lite / NatureBench Lite evaluation & reproduction scripts.

CITE — BIBTEX
@misc{openrsi2026frontisma1,
  title  = {Frontis-MA1: Training an AI4AI Model towards Recursive Self-Improvement in ML Engineering},
  author = {{OpenRSI Team}},
  year   = {2026},
  url    = {https://github.com/FrontisAI/OpenRSI},
  note   = {Preprint}
}