weights

Unsloth has released Multi-Token Prediction (MTP) GGUF weights for Gemma 4 models (31B, 26B-A4B, 12B) in Q8, F16, and BF16 precisions, available on Hugging Face.

Nvidia's Nemotron series of AI models is fully open source, with benchmarks, GitHub repos, data, and weights available, performing competitively with NVFP4 benchmarks only 1% away.

A creative dialogue explores the idea that large language models are fundamentally just matrices of weights, challenging notions of understanding and sentience.

Ideogram 4.0 is released as the best open image model, available for download, fine-tuning, and on all Ideogram plans and the API.

Asking about the percentage of weight changes between Opus 4.7 and Opus 4.8.