tools
updated every night
TRIBE
launchlate March 2026
powered byTRIBE v2
goblin vibe check:
for people building adaptive difficulty or accessibility features who want to predict how players actually process your game stimuli
trimodal foundation model for predicting whole-brain responses to video, audio, and text and enabling large-scale virtual neuroscience experiments.
Unified trimodal brain-response predictionRoughly 70x voxel-resolution jump over the prior systemZero-shot generalization to unseen subjectsTurns months of lab work into fast in-silico experiments
key features
Unified trimodal brain-response predictionRoughly 70x voxel-resolution jump over the prior systemZero-shot generalization to unseen subjectsTurns months of lab work into fast in-silico experiments
spec & usage
Combines a transformer backbone with text and video encoders such as LLaMA 3.2 and V-JEPA2
Trained on more than 1,000 hours of fMRI recordings from 700+ volunteers
Released under CC BY-NC for research use with code and weights available to academic teams
scope:
dataresearchopen-sourcefreemultimodal
launchlate March 2026