What Fits (Into Few Tokens) Doesn't Overfit: Compression and Generalization in ML Research Agents
Researchers have investigated why machine learning, particularly when driven by large language models (LLMs), exhibits surprisingly little overfitting despite adaptive benchmark use. Their study on LLM-driven research agents suggests that successful ML strategies are highly compressible. Experiments with output and input compression, using short prompts and one-bit feedback, demonstrated that these bottlenecks minimally impacted performance across various datasets, supporting the idea that effective strategies occupy a low-complexity region of strategy space. AI
IMPACT Suggests that the inherent compressibility of successful ML strategies may explain the observed lack of overfitting in benchmark-driven ML.