Hugging Face releases Carbon-A: 1.2B model proposes 566M gene candidates across 22,617 species
lvwerra · x · 2026-10-13
Hugging Face Bio unveiled Carbon-A, a 1.2B-parameter foundation model for genome annotation across the eukaryotic tree of life, open-sourced under Apache-2.0.
Key numbers:
- Covers 22,617 species and 48,167 genome assemblies — 11× more species than its RefSeq training corpus
- Proposed 566M gene candidates (green = new annotations, blue = existing database annotations)
- 0.944 nucleotide-level F1 across 42 benchmarks
- Supported experimentally by long-read RNA sequencing
The team frames it as open AI infrastructure for biology; model, database and paper are all linked.
More from Research
- Microsoft's AutoEnvScaling lets agents build their own RL environments, lifting Qwen3 Terminal-Bench from 41.1 to 53.3 — omarsar0 · 2026-10-15
- Microsoft open-sources ML tool that corrects weather model bias, extending reliable forecasts to 2-6 weeks — tomwarren · 2026-10-15
- HNC paper: shifting ViT weights 1.3% keeps predictions but scrambles representations — fatihdin4en · 2026-10-15
- LLM Tool-Call Judges Reject Correct Values Without References: The Derivation Gap — Linhai Ma · 2026-10-15
- Work photos are AI's first mile: if capture is wrong, no model can recover — arnabdotorg · 2026-10-15
- Decoding NCM's latents reveals coarse-to-fine text generation trajectory — Azaliamirh · 2026-10-15