Hugging Face releases Carbon-X, an open-source model built on Qwen3-14B that lets scientists chat with DNA
_lewtun · x · 2026-10-22
Hugging Face launched Carbon-X, billed as a first-of-its-kind open-source model combining DNA analysis and generation with plain language — researchers can ask whether a mutation raises disease risk and get predictions plus explanations, or generate new sequences to aid gene therapy design.
Key facts: it was built by training Qwen3-14B on genomic sequences from humans, other mammals, plants, and fungi, in collaboration with Zhongguancun Academy (China) and University of Naples Federico II (Italy). The release comes as Nvidia's $13B acquisition of Hugging Face, agreed in September, is pending.
Related event: Hugging Face Open-Sources Carbon-X, a 14B DNA-Language Model(20 posts)→
More from Research
- Meta's MemoProBench: frontier models score ≤15% on memory-driven proactive assistance — RulinShao · 2026-10-23
- Clarifying decision-model routing: routing activations across models, not just classifying — ariG23498 · 2026-10-23
- Dev says he built an LLM classifier-head router for MoE back in August 2025 — officialKrishD · 2026-10-23
- Late Interaction releases a short technical report on its late-interaction retrieval models — lateinteraction · 2026-10-23
- New study finds AI chatbots struggle to understand cultural context — Cohere_Labs · 2026-10-23
- Axiom Math open-sources MathEconLib, an econ theory library built with leading economists — BenBlaiszik · 2026-10-23