Inverting Foundation Models of Brain Function with Simulation-Based Inference

TL;DR

This paper demonstrates that foundation models of brain activity can be inverted to recover stimuli and their properties from synthetic brain data using simulation-based inference, enabling decoding and inverse design.

Contribution

It introduces a method to invert brain models with simulation-based inference, allowing recovery of stimuli from neural activity and using LLMs as controllable stimulus generators.

Findings

Stimuli properties can be recovered from predicted brain maps.

LLMs can generate controllable stimuli for simulated experiments.

The approach validates neural encoding quality.

Abstract

Foundation models of brain activity promise a new frontier for in silico neuroscience by emulating neural responses to complex stimuli across tasks and modalities. A natural next step is to ask whether these models can also be used in reverse. Can we recover a stimulus or its properties from synthetic brain activity? We study this question in a proof-of-concept setting using TRIBEv2. We pair the brain emulator with large language models (LLMs) that generate news headlines from linguistic parameters such as valence, arousal, and dominance. We then use simulation-based inference to learn a probabilistic mapping from brain maps to latent stimulus parameters. Our results show that these parameters can be recovered from predicted brain maps, validating the quality of neural encodings. They also show that LLMs can serve as controllable stimulus generators for simulated experiments. Together, these findings provide a step toward decoding and inverse design with foundation brain models.