arXiv:2605.25183v2 Announce Type: replace 
Abstract: Knowledge graph (KG) is an abstraction that can be extracted from text corpora and used for in-depth reasoning. Prior work has leveraged KGs to fine-tune language models (LMs), enabling domain-specific superintelligence. In this work, we explore whether KG-driven in-depth reasoning capabilities can emerge in neuroscience using only information contained within a single authoritative textbook. The central hypothesis is that structured knowledge, when distilled into a high-quality KG and converted into KG-grounded question-answer (QA) supervision, is sufficient to produce expert-level reasoning through a fine-tuned LM that surpasses large language models (LLMs) in accuracy, while employing orders of magnitude fewer parameters. We construct a textbook-derived KG via a dual-LLM validation pipeline, expand it with a masked LM trained on the KG topology, generate multi-hop QA items, which include QA pairs and reasoning traces, to fine-tune an LM exclusively on KG-derived supervision, and apply reinforcement learning using path-derived KG signals as implicit reward models. Our results demonstrate that deep, mechanistic neuroscience understanding can be induced in the model without reliance on large, heterogeneous web-scale corpora. The KG-based synthetic neuroscience curriculum that readers can quiz themselves on, and the fine-tuned LM, are available at the following GitHub location: https://kg-bottom-up-superintelligence.github.io/neuro-bench.

أظهرت دراسة حديثة إمكانية استخدام الرسوم البيانية المعرفية (KG) لتعزيز التفكير الخبير في علم الأعصاب من خلال الاستفادة من المعلومات الموجودة في كتاب نصي موثوق واحد. تفترض الدراسة أن الرسم البياني المعرفي المنظم جيدًا يمكن أن يمكّن نماذج اللغة (LM) المعدلة من تحقيق قدرات تفكير متفوقة، متجاوزة النماذج الأكبر في الدقة مع استخدام عدد أقل بكثير من المعلمات.

Un estudio reciente ha explorado el potencial de los grafos de conocimiento (KG) para mejorar el razonamiento de nivel experto en neurociencia utilizando información de un solo libro de texto autorizado. La investigación postula que un KG bien estructurado puede permitir que los modelos de lenguaje (LM) ajustados logren capacidades de razonamiento superiores, superando a modelos más grandes en precisión mientras utilizan significativamente menos parámetros.

Une étude récente a exploré le potentiel des graphes de connaissances (KG) pour améliorer le raisonnement de niveau expert en neurosciences en utilisant des informations d'un seul manuel autoritaire. La recherche postule qu'un KG bien structuré peut permettre aux modèles de langage (LM) ajustés d'atteindre des capacités de raisonnement supérieures, dépassant les modèles plus grands en précision tout en utilisant significativement moins de paramètres.

A recent study has explored the potential of knowledge graphs (KGs) to enhance expert-level reasoning in neuroscience by utilizing information from a single authoritative textbook. The research posits that a well-structured KG can enable fine-tuned language models (LMs) to achieve superior reasoning capabilities, surpassing larger models in accuracy while using significantly fewer parameters.