# MMPostTrainBench: Benchmarking Autonomous Research for Multimodal Post-Training

arXiv: https://arxiv.org/abs/2610.05398 · HTML: https://mlsanity.com/paper/2610.05398/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 27 of its 28 references that are in the map.
- First posted on arXiv 4 Oct 2026.

## Builds on (most cited first, 27 of 27)

- [The AI Scientist: Towards Fully Automated Open-Ended Scientific Discovery](https://mlsanity.com/paper/2408.06292/) (2024)
- [MMMU-Pro: A More Robust Multi-discipline Multimodal Understanding Benchmark](https://mlsanity.com/paper/2409.02813/) (2024)
- [Qwen3-Omni Technical Report](https://mlsanity.com/paper/2509.17765/) (2025)
- [MLE-bench: Evaluating Machine Learning Agents on Machine Learning Engineering](https://mlsanity.com/paper/2410.07095/) (2024)
- [MLAgentBench: Evaluating Language Agents on Machine Learning Experimentation](https://mlsanity.com/paper/2310.03302/) (2023)
- [MMAU: A Massive Multi-Task Audio Understanding and Reasoning Benchmark](https://mlsanity.com/paper/2410.19168/) (2024)
- [Video-MMMU: Evaluating Knowledge Acquisition from Multi-Discipline Professional Videos](https://mlsanity.com/paper/2501.13826/) (2025)
- [AIDE: AI-Driven Exploration in the Space of Code](https://mlsanity.com/paper/2502.13138/) (2025)
- [SWE-bench Multimodal: Do AI Systems Generalize to Visual Software Domains?](https://mlsanity.com/paper/2410.03859/) (2024)
- [RE-Bench: Evaluating frontier AI R&D capabilities of language model agents against human experts](https://mlsanity.com/paper/2411.15114/) (2024)
- [MME-CoT: Benchmarking Chain-of-Thought in Large Multimodal Models for Reasoning Quality, Robustness, and Efficiency](https://mlsanity.com/paper/2502.09621/) (2025)
- [MMAR: A Challenging Benchmark for Deep Reasoning in Speech, Audio, Music, and Their Mix](https://mlsanity.com/paper/2505.13032/) (2025)
- [PostTrainBench: Can LLM Agents Automate LLM Post-Training?](https://mlsanity.com/paper/2603.08640/) (2026)
- [OmniVideoBench: Towards Audio-Visual Understanding Evaluation for Omni MLLMs](https://mlsanity.com/paper/2510.10689/) (2025)
- [Video-MME-v2: Towards the Next Stage in Benchmarks for Comprehensive Video Understanding](https://mlsanity.com/paper/2604.05015/) (2026)
- [Toward Generalist Autonomous Research via Hypothesis-Tree Refinement](https://mlsanity.com/paper/2606.11926/) (2026)
- [RSIBench-Data: Benchmarking Data-Centric Research for Recursive Self-Improvement](https://mlsanity.com/paper/2607.25886/) (2026)
- [Dream-RSI: Recursive Self-Improvement through Evolving Worlds](https://mlsanity.com/paper/2609.14858/) (2026)
- [JointAVBench: A Benchmark for Joint Audio-Visual Reasoning Evaluation](https://mlsanity.com/paper/2512.12772/) (2025)
- [AIBuildAI: An AI Agent for Automatically Building AI Models](https://mlsanity.com/paper/2604.14455/) (2026)
- [HarnessDev: Can LLMs Create and Evolve Their Own Agent Harness?](https://mlsanity.com/paper/2609.01437/) (2026)
- [ModularRSI: Modular and Generalizable Recursive Harness Self-Improvement](https://mlsanity.com/paper/2609.14857/) (2026)
- [RSIAgent: Autonomous Exploration for Recursive Self-improvement in New Environments](https://mlsanity.com/paper/2609.15364/) (2026)
- [S3Gym: Can LLMs Turn Self-Testing and Self-Judging into Self-Improvement?](https://mlsanity.com/paper/2608.31100/) (2026)
- [Aspire: Can Models Self-Evolve from Vague Goals?](https://mlsanity.com/paper/2608.31111/) (2026)
- [AIBuildAI-2: A Knowledge-Enhanced Agent for Automatically Building AI Models](https://mlsanity.com/paper/2605.27873/) (2026)
- [Phi-Bench: Can Large Language Models Engineer the Infrastructure That Powers Them?](https://mlsanity.com/paper/2609.10226/) (2026)

## Abstract (arXiv)

Autonomous research seeks sustained model improvements through iterative experimentation and feedback. LLM agents show promise in automating machine learning and language-model post-training, but their ability to sustain multimodal improvement remains unclear. We introduce MMPostTrainBench, a benchmark spanning eight tasks in image, audio, video, and joint audio-video understanding and image-grounded software repair. Agents operate from a common base model within fixed budgets, using development feedback before independent evaluation of their submitted models. Evaluation covers target and non-target model outcomes, iterative model improvement and selection, and research integrity. Across all eight tasks, 52.1% of model--task means fall below the base, and evaluated submissions also exhibit non-target regressions. Model performance does not consistently improve across research iterations, and agents do not reliably select the best evaluated candidate for submission; final submissions trail that candidate by up to 5.38 percentage points. Extending autonomous research from text-only to multimodal tasks introduces additional sources of error in perception, cross-modal alignment, and temporal grounding. The observed regressions and selection gaps highlight the need to balance targeted improvements with non-target capability preservation and to retain gains across research iterations. These requirements motivate MMResearch, a multimodal research framework that connects media-grounded evidence to hypotheses and interventions, carries findings across rounds through hierarchical memory, and retains candidates using development evaluation. Added to existing code-agent runtimes, it improves submitted-model accuracy by up to 7.75 percentage points for Claude Opus 4.8 with Claude Code and 2.33 points for GPT-5.6-sol with Codex.