Self-Play for Language Models

Latest papers

8 papers0 papers in the last four weeks, with none the four weeks before. 0.0% of all new papers.

Toward LLM Self-Play Under Data Scarcity: Effectiveness, Generalizability, and Trustworthiness
Nov 6, 2026Zhiyin Yu, Yuchen Mou, Bingzhe Wu +3Self-Play for Language Models Language Model Self-Improvement
Peking University, Beijing, China · Shanghai Artificial Intelligence Laboratory, Shanghai, China · the National University of Singapore, Singapore +2
Test-Time Self-Play: Scaling without a Verifier
Nov 5, 2026Akira Yoshiyama, Akmal Ashirmatov, Jonas Hübotter +2Self-Play for Language Models Test-Time Scaling
ETH Zurich & Tufa Labs · ETH Zurich
Learning What to Practice: Diagnosis-Guided Self-Evolution for Language Models
Sep 1, 2026Xincheng Wei, Yifan Ding, Fucheng Xiong +6Self-Play for Language Models Self-Training for Language Models
The Chinese University of Hong Kong, Shenzhen · Meituan, LongCat Team · Peking University +2
SearchMaster: Grounded and Regulated Self-Play for Search Agents
Aug 3, 2026Wentao Tan, Qiong Cao, Jiaqi Wang +1Multi-Hop QA Self-Play for Language Models
JD Future Academy, Beijing
Audio-Zero: Label-Free Self-Evolution for Fine-Grained Audio Reasoning
Jul 22, 2026Siqian Tong, Xuan Li, Chaozhuo Li +5Self-Play for Language Models Audio Reasoning
Institute of Acoustics, Chinese Academy of Sciences · University of Chinese Academy of Sciences · 3Beijing Academy of Artificial Intelligence +3
Hallucination Self-Play: Bootstrapping Reinforced Detector via Evolved Generator
Jul 8, 2026Shiping Yang, Shining Liang, Weihao Liu +4Self-Play for Language Models LLM Hallucination Detection
Simon Fraser University · Microsoft · Work done during an internship at Microsoft +1
D $^2$ Evo: Dual Difficulty-Aware Self-Evolution for Data-Efficient Reinforcement Learning
May 16, 2026Ru Zhang, Renda Li, Ziyu Ma +4Self-Play for Language Models RL for Language Model Reasoning
Alibaba Group · AMAP, Alibaba Group · BUPT
SELF-EMO: Emotional Self-Evolution from Recognition to Consistent Expression
Apr 20, 2026Shaowei Zhang, Faqiang Qian, Yan Chen +5RL for Language Models Self-Play for Language Models
1SenseTime · 2Tsinghua University · 3Shanghai Jiao Tong University +1

Nov 6, 2026J/K move · Enter open · S save

Toward LLM Self-Play Under Data Scarcity: Effectiveness, Generalizability, and Trustworthiness

Zhiyin Yu, Yuchen Mou, Bingzhe Wu, Zhonghai Wu+2

Peking University, Beijing, China · Shanghai Artificial Intelligence Laboratory, Shanghai, China · the National University of Singapore, Singapore+2

Self-Play for Language Models Language Model Self-Improvement Self-Play RL Synthetic Data Generation for Language Models LLM Reliability

Open paper AskPDF arXiv

The rapid development of large language models (LLMs) has been driven largely by scaling model size and human-curated training data, yet this paradigm is increasingly constrained by data scarcity and diminishing returns from human supervision. Self-play offers a promising pathway for autonomous data generation and improvement, but current frameworks remain unsatisfactory due to a lack of a data-centric perspective, limited cross-domain generalization, and growing trustworthiness and security concerns. In this paper, we argue that real-world LLM self-play systems should achieve effectiveness, generalizability, and trustworthiness under data scarcity. To address these challenges, we outline three blue-sky research directions for future self-play systems: (1) data-centric self-play for effective experience generation, (2) generalizable self-play for open-world and unverifiable tasks, and (3) trustworthy self-play for reliable self-evolution. We further discuss research opportunities and solutions within each direction. This paradigm shift will establish a sustainable and safe AI ecosystem by transforming finite human knowledge into inexhaustible high-quality experience through self-play architectures.