← Back to People
Chengsong Huang is a CSE Ph.D. candidate at Washington University in St. Louis researching self-improving LLMs and cost-efficient use of large models.
1
Mentions
1
Articles
1
Outlets
Associated Topics
Companies Covered
Coverage Patterns
How media typically covers Chengsong Huang
Article Types
- research1
Media Angles
- technical1
Narrative Framing
- breakthrough1
Associated AI Models
Qwen3-4B-Base1
Articles
Most recent firstArticles Written
Chengsong Huang as author
R-Zero: Self-Evolving Reasoning LLM From Zero Data
arXivresearchpositiveAug 29, 2025
R-Zero is a fully autonomous framework that trains LLMs without human-curated data by having a Challenger model propose tasks and a Solver model solve them in a co-evolving curriculum, achieving +6.49 improvement on math reasoning for Qwen3-4B-Base.
“Author of "R-Zero: Self-Evolving Reasoning LLM From Zero Data" in arXiv”
LLMsAI ResearchReinforcement LearningFoundation Models