# From Procedural Skills to Strategy Genes: Towards Experience-Driven Test-Time Evolution

arXiv: https://arxiv.org/abs/2604.15097 · HTML: https://mlsanity.com/paper/2604.15097/

## Key facts

- Cited by 9 papers in the map (data through 14 Oct 2026).
- Builds on 20 of its 24 references that are in the map.
- First posted on arXiv 16 Apr 2026.

## Builds on (most cited first, 20 of 20)

- [Reflexion: Language Agents with Verbal Reinforcement Learning](https://mlsanity.com/paper/2303.11366/) (2023)
- [Self-Refine: Iterative Refinement with Self-Feedback](https://mlsanity.com/paper/2303.17651/) (2023)
- [A Survey on Large Language Model based Autonomous Agents](https://mlsanity.com/paper/2308.11432/) (2023)
- [Voyager: An Open-Ended Embodied Agent with Large Language Models](https://mlsanity.com/paper/2305.16291/) (2023)
- [Teaching Large Language Models to Self-Debug](https://mlsanity.com/paper/2304.05128/) (2023)
- [ExpeL: LLM Agents Are Experiential Learners](https://mlsanity.com/paper/2308.10144/) (2023)
- [CRITIC: Large Language Models Can Self-Correct with Tool-Interactive Critiquing](https://mlsanity.com/paper/2305.11738/) (2023)
- [A Survey on the Memory Mechanism of Large Language Model based Agents](https://mlsanity.com/paper/2404.13501/) (2024)
- [SkillsBench: Benchmarking How Well Agent Skills Work Across Diverse Tasks](https://mlsanity.com/paper/2602.12670/) (2026)
- [MemSkill: Learning and Evolving Memory Skills for Self-Evolving Agents](https://mlsanity.com/paper/2602.02474/) (2026)
- [Evo-Memory: Benchmarking LLM Agent Test-time Learning with Self-Evolving Memory](https://mlsanity.com/paper/2511.20857/) (2025)
- [Agent Skills for Large Language Models: Architecture, Acquisition, Security, and the Path Forward](https://mlsanity.com/paper/2602.12430/) (2026)
- [Memp: Exploring Agent Procedural Memory](https://mlsanity.com/paper/2508.06433/) (2025)
- [MemEvolve: Meta-Evolution of Agent Memory Systems](https://mlsanity.com/paper/2512.18746/) (2025)
- [Memento-Skills: Let Agents Design Agents](https://mlsanity.com/paper/2603.18743/) (2026)
- [Probing the Critical Point (CritPt) of AI Reasoning: a Frontier Physics Research Benchmark](https://mlsanity.com/paper/2509.26574/) (2025)
- [Evaluating AGENTS.md: Are Repository-Level Context Files Helpful for Coding Agents?](https://mlsanity.com/paper/2602.11988/) (2026)
- [TAME: A Trustworthy Test-Time Evolution of Agent Memory with Systematic Benchmarking](https://mlsanity.com/paper/2602.03224/) (2026)
- [UMEM: Unified Memory Extraction and Management Framework for Generalizable Memory](https://mlsanity.com/paper/2602.10652/) (2026)
- [CoWork-X: Experience-Optimized Co-Evolution for Multi-Agent Collaboration System](https://mlsanity.com/paper/2602.05004/) (2026)

## Built on by (most cited first, 9 of 9)

- [SkillsVote: Lifecycle Governance of Agent Skills from Collection, Recommendation to Evolution](https://mlsanity.com/paper/2605.18401/) (2026)
- [MOSS: Self-Evolution through Source-Level Rewriting in Autonomous Agent Systems](https://mlsanity.com/paper/2605.22794/) (2026)
- [Library Drift: Diagnosing and Fixing a Silent Failure Mode in Self-Evolving LLM Skill Libraries](https://mlsanity.com/paper/2605.19576/) (2026)
- [Ratchet: How Reliable Must an LLM Judge Be to Retire a Skill?](https://mlsanity.com/paper/2605.22148/) (2026)
- [Living-Harness Is an Interactive-Agent Evolver](https://mlsanity.com/paper/2607.26598/) (2026)
- [Who Grades the Grader? Co-Evolving Evaluation Metrics and Skills for Self-Improving LLM Agents](https://mlsanity.com/paper/2607.12790/) (2026)
- [Behind EvoMap: Characterizing a Self-Evolving Agent-to-Agent Collaboration Network](https://mlsanity.com/paper/2605.25815/) (2026)
- [LongWoF-Bench: Evaluating EvoMap Genes for Verifiable Long-Workflow Tasks](https://mlsanity.com/paper/2608.23200/) (2026)
- [Who Verifies the Verifier? Co-Evolving Inspectable Graders with Self-Improving Agents](https://mlsanity.com/paper/2610.11464/) (2026)

## Abstract (arXiv)

This beta technical report asks how reusable experience should be represented so that it can function as effective test-time control and as a substrate for iterative evolution. We study this question in 4.590 controlled trials across 45 scientific code-solving scenarios. We find that documentation-oriented Skill packages provide unstable control: their useful signal is sparse, and expanding a compact experience object into a fuller documentation package often fails to help and can degrade the overall average. We further show that representation itself is a first-order factor. A compact Gene representation yields the strongest overall average, remains competitive under substantial structural perturbations, and outperforms matched-budget Skill fragments, while reattaching documentation-oriented material usually weakens rather than improves it. Beyond one-shot control, we show that Gene is also a better carrier for iterative experience accumulation: attached failure history is more effective in Gene than in Skill or freeform text, editable structure matters beyond content alone, and failure information is most useful when distilled into compact warnings rather than naively appended. On CritPt, gene-evolved systems improve over their paired base models from 9.1% to 18.57% and from 17.7% to 27.14%. These results suggest that the core problem in experience reuse is not how to supply more experience, but how to encode experience as a compact, control-oriented, evolution-ready object.