EvoNarrator: Modeling Scientific Evolution for Feasible Hypothesis Generation

ACL ARR 2026 January Submission9969 Authors

Keywords: Automated Hypothesis Generation; Scientific Idea Generation; LLM Agents

Abstract:

Scientific discovery evolution does not emerge in isolation but stems from the structural deepening and recombination of existing functionalities.
However, current automated hypothesis generation methods, constrained by the statistical co-occurrence nature of Large Language Models (LLMs), lack perception of temporal causality and the "evolutionary patterns" inherent in scientific development.
Consequently, they often yield superficial combinations that are logically infeasible. To address this, we propose EvoNarrator, a framework for hypothesis generation based on evolutionary narratives.
We first extract structured P-M-L-F (Problem, Method, Limitation, Future Work) quadruples from citation networks. Subsequently, we introduce the SocketMatch mechanism, which eliminates logical disconnects between methods and problems by assessing their deep semantic compatibility. Finally, utilizing three evolutionary patterns, Divergence, and Convergence—we constrain the generation process within historically logical derivation paths.
Double-blind expert reviews confirm EvoNarrator's superior logicality (4.80/5.00) and predictive foresight via hindcasting experiments. Crucially, ablation studies reveal that integrating evolutionary patterns shifts the paradigm from conservative incrementalism to theoretically grounded structural innovation. The code is
available at https://anonymous.4open.science/r/EvoNarrator-A663.

Paper Type: Long

Research Area: Natural Language Generation

Research Area Keywords: Generation, Information Retrieval and Text Mining, semantic textual similarity, Summarization

Languages Studied: English

Submission Number: 9969