OpenAI: promises to unearth the Machine God™, releases another product feature you’ll forget exists in a week Anthropic: "here's an alignment blogpost and oh, our model is 0.1% better at coding now!" DeepMind: models can solve protein folding and generate new realities, but still no one uses Gemini Meta: simulating whole human brains for better Insta Reels recommendations
AI at Meta
AI at Meta11.8. klo 19.20
🏆 We're thrilled to announce that Meta FAIR’s Brain & AI team won 1st place at the prestigious Algonauts 2025 brain modeling competition. Their 1B parameter model, TRIBE (Trimodal Brain Encoder), is the first deep neural network trained to predict brain responses to stimuli across multiple modalities, cortical areas, and individuals. The approach combines pretrained representations of several foundational models from Meta – text (Llama 3.2), audio (Wav2Vec2-BERT from Seamless) and video (V-JEPA 2) – to predict a very large amount (80 hours per subject) of spatio-temporal fMRI brain responses to movies acquired by the Courtois NeuroMod project Download the code: Read the paper: Learn about the challenge: Download the data:
362,93K