Prewire · case file · public copy of the record
FILE P392 PREWIRE receipts archive the live wire
CATCH #2092 DETECTED OCT 15, 2026 22:20 UTC
THE CATCHscore 10/10
EXHIBIT 2092 · FLASH
Oct 15, 2026 22:20 UTC huggingface nvidia
Nvidia released new Nemotron 3.5 Embed models, 1B and 8B, in BF16 precision, on HuggingFace.

READY TO POST:

nvidia just dropped nemotron 3.5 embed models (1b and 8b) on huggingface. looks like a new embedding family.

new nemotron 3.5 embed models from nvidia just appeared on hf! interesting to see them in bf16 already.

EVIDENCE — the receipts
New HuggingFace repo(s) from nvidia:

  + https://huggingface.co/nvidia/Nemotron-3.5-Embed-1B-BF16
  + https://huggingface.co/nvidia/Nemotron-3.5-Embed-8B-BF16

Labs often push model weights to HF (sometimes gated) before the announcement. Verify whether it's a model, dataset, or space.

caught by the 20-minute sweep · the live wire · all receipts

THE WIRE RECEIPTS ACCESS @VEDOLOS TERMS PRIVACY break the story before it breaks