# SERPO: Self-Evolving Rubric Policy Optimization for Open-Ended Test-Time Reinforcement Learning

arXiv: https://arxiv.org/abs/2607.26873 · HTML: https://mlsanity.com/paper/2607.26873/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Placed in Self-evolving LLM agents, one of 13 directions in Synthetic data and self-improvement for LLMs.
- Builds on 31 of its 37 references that are in the map.
- First posted on arXiv 29 Jul 2026.

## Placed in

- Synthetic data and self-improvement for LLMs / [Self-evolving LLM agents](https://mlsanity.com/synth/self-evolving-llm-agents/)
- Deep reinforcement learning methods / [RL training of language and agent models](https://mlsanity.com/deeprl/rl-training-of-language-and-agent-models/)

## Builds on (most cited first, 30 of 31)

- [Proximal Policy Optimization Algorithms](https://mlsanity.com/paper/1707.06347/) (2017)
- [DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models](https://mlsanity.com/paper/2402.03300/) (2024)
- [Reflexion: Language Agents with Verbal Reinforcement Learning](https://mlsanity.com/paper/2303.11366/) (2023)
- [GPQA: A Graduate-Level Google-Proof Q&A Benchmark](https://mlsanity.com/paper/2311.12022/) (2023)
- [G-Eval: NLG Evaluation using GPT-4 with Better Human Alignment](https://mlsanity.com/paper/2303.16634/) (2023)
- [What Disease does this Patient Have? A Large-scale Open Domain Question Answering Dataset from Medical Exams](https://mlsanity.com/paper/2009.13081/) (2020)
- [HealthBench: Evaluating Large Language Models Towards Improved Human Health](https://mlsanity.com/paper/2505.08775/) (2025)
- [Agentic Context Engineering: Evolving Contexts for Self-Improving Language Models](https://mlsanity.com/paper/2510.04618/) (2025)
- [Rubrics as Rewards: Reinforcement Learning Beyond Verifiable Domains](https://mlsanity.com/paper/2507.17746/) (2025)
- [FLASK: Fine-grained Language Model Evaluation based on Alignment Skill Sets](https://mlsanity.com/paper/2307.10928/) (2023)
- [DR Tulu: Reinforcement Learning with Evolving Rubrics for Deep Research](https://mlsanity.com/paper/2511.19399/) (2025)
- [RubricHub: A Comprehensive and Highly Discriminative Rubric Dataset via Automated Coarse-to-Fine Generation](https://mlsanity.com/paper/2601.08430/) (2026)
- [ResearchQA: Evaluating Scholarly Question Answering at Scale Across 75 Fields with Survey-Mined Questions and Rubrics](https://mlsanity.com/paper/2509.00496/) (2025)
- [LLMEval-Med: A Real-world Clinical Benchmark for Medical LLMs with Physician Validation](https://mlsanity.com/paper/2506.04078/) (2025)
- [LLM-as-a-Verifier: A General-Purpose Verification Framework](https://mlsanity.com/paper/2607.05391/) (2026)
- [X-Eval: Generalizable Multi-aspect Text Evaluation via Augmented Instruction Tuning with Auxiliary Evaluation Aspects](https://mlsanity.com/paper/2311.08788/) (2023)
- [Reinforcing Chain-of-Thought Reasoning with Self-Evolving Rubrics](https://mlsanity.com/paper/2602.10885/) (2026)
- [TTCS: Test-Time Curriculum Synthesis for Self-Evolving](https://mlsanity.com/paper/2601.22628/) (2026)
- [G-Zero: Self-Play for Open-Ended Generation from Zero Data](https://mlsanity.com/paper/2605.09959/) (2026)
- [Rubric-Guided Self-Distillation: Post-Training Without Rubric Verifiers](https://mlsanity.com/paper/2606.12507/) (2026)
- [Step-wise Rubric Rewards for LLM Reasoning](https://mlsanity.com/paper/2605.17291/) (2026)
- [EvoRubrics: Dynamic Rubrics as Rewards via Adversarial Co-Evolution for LLM Reinforcement Learning](https://mlsanity.com/paper/2606.23038/) (2026)
- [EvoLM: Self-Evolving Language Models through Co-Evolved Discriminative Rubrics](https://mlsanity.com/paper/2605.03871/) (2026)
- [EvoRubric: Self-Evolving Rubric-Driven RL for Open-Ended Generation](https://mlsanity.com/paper/2605.29847/) (2026)
- [Co-Evolving LLM Evaluators and Policies via DynamicRubric](https://mlsanity.com/paper/2607.20083/) (2026)
- [Detecting and Mitigating the Correct-Answer Extinction Window in Test-Time Reinforcement Learning with Majority Voting](https://mlsanity.com/paper/2605.19444/) (2026)
- [SCOPE: Self-Play via Co-Evolving Policies for Open-Ended Tasks](https://mlsanity.com/paper/2605.31433/) (2026)
- [Recursive Self-Evolving Agents via Held-Out Selection](https://mlsanity.com/paper/2606.28374/) (2026)
- [When More Sampling Hurts: The Modal Ceiling and Correlation Ceiling of Test-Time Scaling](https://mlsanity.com/paper/2606.28661/) (2026)
- [Bootstrapping Post-training Signals for Open-ended Tasks via Rubric-based Self-play on Pre-training Text](https://mlsanity.com/paper/2604.20051/) (2026)

## Built on by (most cited first, 2 of 2)

- [Reinforcement Learning with Comparative Evidence for Social Intelligence](https://mlsanity.com/paper/2610.04072/) (2026)
- [Verify Less, Evolve More: Training Idea-Level Critics for Verification-Efficient ML Evolving Agents](https://mlsanity.com/paper/2610.08993/) (2026)

## Abstract (arXiv)

Test-time reinforcement learning (TTRL) enables language models to self-evolve at inference time without labeled feedback. Existing methods rely on answer voting and therefore do not extend naturally to open-ended generation, where valid responses cannot be mapped to a shared canonical answer. Without external reward models or stronger judges, adaptation must instead construct reliable rewards from the model's own outputs. We introduce SERPO (Self-Evolving Rubric Policy Optimization), which replaces answer voting with a closed loop that co-evolves response evidence, query-specific rubrics, and policy parameters. Good-Normal-Bad (G-N-B) response evolution organizes maximally separated rollouts into ordered archives; rubric evolution retains criteria that discriminate these archives; probabilistic criterion scoring converts verdict-token likelihoods into reward signals; and policy evolution optimizes the actor with the resulting signals. New actor rollouts then refresh both the archives and rubrics, closing the three-way evolution loop. Across two model configurations, two in-domain benchmarks, and four OOD benchmarks, SERPO improves HealthBench and ResearchQA by up to 20.63 and 20.31 points over the corresponding base models, raises the six-benchmark macro-average by up to 8.06 points, and supports OOD transfer and continued cross-benchmark evolution.