Cognitive cartography of mammalian brains using meta-analysis of AI experts
Abstract
No individual mind can fully grasp the complexity of modern neuroscientific literature. Where human experts flag, the latest AI models excel. Here we show how frontier large language models can systematically and quantitatively summarise literature-wide neuroscientific knowledge about the cognitive functions and dysfunctions associated with each brain region. Meta-analysis of these AI experts reveals structure-function mappings to which existing meta-analytic frameworks are blind, demonstrated by causal manipulations including lesions and intracranial stimulation. It also unlocks the possibility of extending quantitative literature meta-analysis and decoding of brain maps to other model organisms beyond human. As proof-of-concept, we integrate LLM meta-analysis with species-specific transcriptomics and connectomics in human, macaque, and mouse, discovering that the human molecular circuit for cognition is evolutionarily conserved. Altogether, literature synthesis with AI experts can fundamentally catalyze neuroscientific discovery by overcoming the barrier of data aggregation from heterogeneous studies, finally bringing together a scattered literature to identify latent insights across disparate subfields, modalities, and species.
Related
- From Genes to Brains: Molecular Evolution and Mammalian Brain Cellular Diversity
- Effects of exercise on cognitive function in older adults: a systematic review and meta-analysis of the mediating role of sleep
- Chemical cartography identifies potential mechanistic, diagnostic, and therapeutic opportunities in RAG deficiency
- Age-related decline in directed forgetting: a meta-analytic test of inhibition versus cognitive load accounts.
- Single-cell mass cytometry cartography highlights systemic lupus erythematosus and Sjögren’s disease immune differences
BibTeX
@misc{luppi2026cognitive,
title = {{Cognitive cartography of mammalian brains using meta-analysis of AI experts}},
author = {Andrea I Luppi and Hana Ali and Zhen-Qi I Liu and Filip Milisav and Arianna Sala and Jitka Annen and Alessandro Gozzi and Danilo Bzdok and Bratislav Misic},
howpublished = {bioRxiv},
year = {2026},
month = oct,
doi = {10.64898/2025.12.01.691701},
url = {https://doi.org/10.64898/2025.12.01.691701},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Cognitive cartography of mammalian brains using meta-analysis of AI experts
AU - Andrea I Luppi
AU - Hana Ali
AU - Zhen-Qi I Liu
AU - Filip Milisav
AU - Arianna Sala
AU - Jitka Annen
AU - Alessandro Gozzi
AU - Danilo Bzdok
AU - Bratislav Misic
PY - 2026
DA - 2026/10/19
PB - bioRxiv
DO - 10.64898/2025.12.01.691701
UR - https://doi.org/10.64898/2025.12.01.691701
AB - No individual mind can fully grasp the complexity of modern neuroscientific literature. Where human experts flag, the latest AI models excel. Here we show how frontier large language models can systematically and quantitatively summarise literature-wide neuroscientific knowledge about the cognitive functions and dysfunctions associated with each brain region. Meta-analysis of these AI experts reveals structure-function mappings to which existing meta-analytic frameworks are blind, demonstrated by causal manipulations including lesions and intracranial stimulation. It also unlocks the possibility of extending quantitative literature meta-analysis and decoding of brain maps to other model organisms beyond human. As proof-of-concept, we integrate LLM meta-analysis with species-specific transcriptomics and connectomics in human, macaque, and mouse, discovering that the human molecular circuit for cognition is evolutionarily conserved. Altogether, literature synthesis with AI experts can fundamentally catalyze neuroscientific discovery by overcoming the barrier of data aggregation from heterogeneous studies, finally bringing together a scattered literature to identify latent insights across disparate subfields, modalities, and species.
SN - 2692-8205
ER -