Unlocking LLM Creativity in Science through Analogical Reasoning

2026-05-11

AI Summary

This study addresses the tendency of large language models (LLMs) to produce unoriginal and homogeneous outputs when tackling open-ended scientific problems—a limitation often attributed to mode collapse. To mitigate this, the work introduces, for the first time, a systematic integration of analogical reasoning into LLM-driven scientific discovery. Specifically, it proposes an analogy generation and transfer method grounded in cross-domain relational structure matching, which guides the model toward more creative and diverse solutions. Experimental results across four biomedical tasks demonstrate substantial improvements: solution diversity increases by 90–173%, over 50% of generated outputs exhibit novelty, distributional metrics improve by nearly 13-fold, and state-of-the-art performance is achieved in oligonucleotide property prediction.

Abstract

Autonomous science promises to augment scientific discovery, particularly in complex fields like biomedicine. However, this requires AI systems that can consistently generate novel and diverse solutions to open-ended problems. We evaluate LLMs on the task of open-ended solution generation and quantify their tendency to mode collapse into low-diversity generations. To mitigate this mode collapse, we introduce analogical reasoning (AR) as a new approach to solution generation. AR generates analogies to cross-domain problems based on shared relational structure, then uses those analogies to search for novel solutions. Compared to baselines, AR discovers significantly more diverse generations (improving solution diversity metrics by 90-173%), generates novel solutions over 50% of the time (compared to as little as 1.6% for baselines), and produces high-quality analogies. To validate the real-world feasibility of AR, we implement AR-generated solutions across four biomedical problems, yielding consistent quantitative gains. AR-generated approaches achieve a nearly 13-fold improvement on distributional metrics for perturbation effect prediction, outperform all baselines on AUPRC when predicting cell-cell communication, infer brain region interactions with a high Spearman correlation ($ρ$=0.729) to published methods, and establish state-of-the-art performance on 2 datasets for oligonucleotide property prediction. The novel and diverse solutions produced by AR can be used to augment the search space of existing solution generation methods.

Problem

• mode collapse
• solution diversity
• open-ended problem solving
• scientific creativity
• large language models

Innovation

• analogical reasoning
• large language models
• solution diversity
• mode collapse
• scientific discovery