Policy Induction: Predicting Startup Success via Explainable Memory-Augmented In-Context Learning
This addresses the high-risk, data-scarce challenge of early-stage startup investment for investors and domain experts, offering an incremental improvement in interpretability and efficiency over traditional methods.
The paper tackles the problem of predicting startup success with scarce data by proposing a transparent and data-efficient framework using memory-augmented LLMs with in-context learning, achieving over 20x more precision than random chance and 7.1x more than top-tier VC firms.
Early-stage startup investment is a high-risk endeavor characterized by scarce data and uncertain outcomes. Traditional machine learning approaches often require large, labeled datasets and extensive fine-tuning, yet remain opaque and difficult for domain experts to interpret or improve. In this paper, we propose a transparent and data-efficient investment decision framework powered by memory-augmented large language models (LLMs) using in-context learning (ICL). Central to our method is a natural language policy embedded directly into the LLM prompt, enabling the model to apply explicit reasoning patterns and allowing human experts to easily interpret, audit, and iteratively refine the logic. We introduce a lightweight training process that combines few-shot learning with an in-context learning loop, enabling the LLM to update its decision policy iteratively based on structured feedback. With only minimal supervision and no gradient-based optimization, our system predicts startup success far more accurately than existing benchmarks. It is over 20x more precise than random chance, which succeeds 1.9% of the time. It is also 7.1x more precise than the typical 5.6% success rate of top-tier venture capital (VC) firms.