CoSPlay: Cooperative Self-Play at Test-Time with Self-Generated Code and Unit Test

Hugging Face Daily Papers Papers
self-play code-generation unit-test test-time-scaling rlvr llm

Summary

CoSPlay is a training-free framework that jointly improves code generation and unit test quality through cooperative self-play, achieving competitive performance without ground-truth unit tests.

Recently, Reinforcement Learning with Verifiable Rewards (RLVR) and Test-Time Scaling (TTS) have advanced LLM code generation through executable verification. Yet Ground-Truth Unit Tests (GT UTs) remain a bottleneck: SOTA RLVR methods require them for costly training, while existing TTS methods lose competitiveness without them. This motivates GT-free TTS, where existing methods directly use self-generated UTs to refine and select code candidates. Yet such UTs are often noisy or spuriously coupled with wrong code, and UT quality in turn cannot be validated without reliable code. The key challenge is therefore to jointly improve both. To this end, we present CoSPlay, a GT-free, training-free framework that jointly improves codes and UTs through cooperative self-play. It first explores diverse solution ideas and identifies their potential failure modes to produce discriminative UT ideas. It then uses bidirectional pass-count signals from the Code-UT execution matrix to iteratively prune or fix weak codes and refresh or replace unreliable UTs, letting the two pools co-evolve. Finally, when multiple codes remain tied at the highest pass count, it picks the final code from the largest output-consensus cluster, since correct codes agree on the same inputs while wrong codes diverge. Experiments on four challenging benchmarks show that CoSPlay on Qwen2.5-7B-Instruct improves average BoN from 22.1% to 33.2% and UT accuracy from 14.6% to 78.3%, matching or surpassing the RLVR model CURE-7B. When applied to CURE-7B, it further improves BoN by 5.7%. CoSPlay also generalizes across diverse backbones and outperforms GT-free TTS baselines under comparable token budgets, with continued gains as the budget scales up. These results suggest a scalable inference strategy for competitive code generation without any GT data.
Original Article
View Cached Full Text

Cached at: 05/26/26, 06:45 PM

Paper page - CoSPlay: Cooperative Self-Play at Test-Time with Self-Generated Code and Unit Test

Source: https://huggingface.co/papers/2605.23491

Abstract

CoSPlay is a GT-free framework that jointly improves code generation and unit test quality through cooperative self-play, achieving competitive performance without ground-truth unit tests.

Recently,Reinforcement Learning with Verifiable Rewards(RLVR) andTest-Time Scaling(TTS) have advanced LLMcode generationthrough executable verification. YetGround-Truth Unit Tests(GT UTs) remain a bottleneck: SOTA RLVR methods require them for costly training, while existing TTS methods lose competitiveness without them. This motivates GT-free TTS, where existing methods directly use self-generated UTs to refine and select code candidates. Yet such UTs are often noisy or spuriously coupled with wrong code, and UT quality in turn cannot be validated without reliable code. The key challenge is therefore to jointly improve both. To this end, we present CoSPlay, a GT-free, training-free framework that jointly improves codes and UTs throughcooperative self-play. It first explores diverse solution ideas and identifies their potential failure modes to produce discriminative UT ideas. It then usesbidirectional pass-count signalsfrom the Code-UT execution matrix to iteratively prune or fix weak codes and refresh or replace unreliable UTs, letting the two pools co-evolve. Finally, when multiple codes remain tied at the highest pass count, it picks the final code from the largestoutput-consensus cluster, since correct codes agree on the same inputs while wrong codes diverge. Experiments on four challenging benchmarks show that CoSPlay onQwen2.5-7B-Instructimproves average BoN from 22.1% to 33.2% and UT accuracy from 14.6% to 78.3%, matching or surpassing the RLVR modelCURE-7B. When applied toCURE-7B, it further improves BoN by 5.7%. CoSPlay also generalizes across diverse backbones and outperforms GT-free TTS baselines under comparable token budgets, with continued gains as the budget scales up. These results suggest a scalable inference strategy for competitivecode generationwithout any GT data.

View arXiv pageView PDFGitHub0Add to collection

Models citing this paper0

No model linking this paper

Cite arxiv.org/abs/2605.23491 in a model README.md to link it from this page.

Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2605.23491 in a dataset README.md to link it from this page.

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2605.23491 in a Space README.md to link it from this page.

Collections including this paper0

No Collection including this paper

Add this paper to acollectionto link it from this page.

Similar Articles

Competitive self-play

OpenAI Blog

OpenAI demonstrates that competitive self-play in simulated 3D robot environments enables AI agents to discover complex physical behaviors like tackling, ducking, and faking without explicit instruction, suggesting self-play will be fundamental to future powerful AI systems.

PlayCoder: Making LLM-Generated GUI Code Playable

Hugging Face Daily Papers

PlayCoder introduces PlayEval benchmark and a multi-agent framework that iteratively repairs LLM-generated GUI applications, achieving up to 20.3% end-to-end playable code.

CodeAlchemy: Synthetic Code Rewriting at Scale

arXiv cs.CL

CodeAlchemy is a synthetic data generation framework that transforms publicly available code into semantically rich training data using five strategies, producing over 500 billion tokens and enabling small models to outperform much larger ones on code benchmarks.