ResearchCodeBench: Benchmarking LLMs on Implementing Novel Machine Learning Research Code
This provides a benchmark for researchers and developers to assess and improve LLMs' capability in generating research code, addressing a specific gap in evaluating their real-world applicability in machine learning innovation.
The paper tackles the problem of evaluating large language models' ability to implement novel machine learning research ideas from recent papers into executable code, finding that even the best models achieve less than 40% success rate, with Gemini-2.5-Pro-Preview performing best at 37.3%.
Large language models (LLMs) have shown promise in transforming machine learning research, yet their capability to faithfully implement novel ideas from recent research papers-ideas unseen during pretraining-remains unclear. We introduce ResearchCodeBench, a benchmark of 212 coding challenges that evaluates LLMs' ability to translate cutting-edge ML contributions from top 2024-2025 research papers into executable code. We assessed 30+ proprietary and open-source LLMs, finding that even the best models correctly implement less than 40% of the code. We find Gemini-2.5-Pro-Preview to perform best at 37.3% success rate, with O3 (High) and O4-mini (High) following behind at 32.3% and 30.8% respectively. We present empirical findings on performance comparison, contamination, and error patterns. By providing a rigorous and community-driven evaluation platform, ResearchCodeBench enables continuous understanding and advancement of LLM-driven innovation in research code generation.