Codes and data for "Emergence of Social Reasoning through Adaptive Evolution of LLMs"
Rattachement africain : jp. Niveau de preuve : code pays fourni par la source.
Le résumé fourni par la source
Code and data for the ALIFE 2026 paper "Emergence of Social Reasoning through Adaptive Evolution of LLMs." Acknowledgment of Upstream Codebase:Our evolutionary computation framework is built upon GENOME (Zhang et al., 2025). We deeply appreciate the original authors for open-sourcing their excellent framework, which serves as the core engine for evolving LoRA adapters in this study. Please see the README.md and THIRD_PARTY_LICENSES for detailed attribution and citation information. Status:This is the initial minimal release; an organized version will be added later on the same record. Repository Structure:For this Zenodo release, the files are distributed in two archives: system.zip: Contains the core evolutionary computation framework, analysis tools, and datasets. experiment_result.zip: Contains the evolved LoRA adapters, state logs, and generated figures reported in the paper. Important License Notice:This repository contains mixed licensing. Please refer to the README.md and THIRD_PARTY_LICENSES for full details. Code: MIT License (Derivative work of GENOME). Result Data (Logs, Figures, CSVs): CC-BY-4.0. Evolved LoRA Adapters (Model Weights): Strictly subject to the Meta Llama 3 Community License and the Gemma Terms of Use. The CC-BY-4.0 license does NOT apply to these derivative model weights. Any downstream use, distribution, or modification of these weights must comply with the original upstream licenses and Acceptable Use Policies.
Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.
Le contrôle bibliographique ouvert
Où se fait cette recherche
-
Nagoya University pays non établi dans la noticeUniversité ou école supérieure
Nagoya University.
Une affiliation ne permet pas de déduire la nationalité d’un auteur.