Anthropic releases 211 scorpion-toxin designs after 17 meet binding criteria

Anthropic announced on October 20 that its internal Claude research model generated 318,484 protein designs targeting scorpion neurotoxins, and that 17 of the 211 designs sent for outside lab testing met the company's binding criteria. Anthropic published the designs and associated data on Hugging Face, but the company said the tests showed binding only, with no evidence any design neutralizes venom or treats a sting, and affordable production remains unresolved.

Anthropic releases 211 scorpion-toxin designs after 17 meet binding criteria Anthropic's Claude generated 318,484 candidates in a 64-hour campaign; lab tests found binding, not evidence that any design neutralizes venom or treats a sting. By RuntimeWire Staff https://runtimewire.com/author/runtimewire-staff · Published Primary source: Anthropic https://www.anthropic.com/research/claude-scorpion-toxin-binder-design Why it matters Claude helped generate toxin-binder candidates that outside labs tested during a 64-hour campaign; 17 met binding criteria. Whether any can neutralize venom or become a safe and practical medicine remains unknown. Affordable production is a separate unresolved challenge. Anthropic says its internal Claude research model https://www.anthropic.com/research/claude-scorpion-toxin-binder-design?ref=runtimewire generated 318,484 protein designs aimed at scorpion neurotoxins, with 17 of the 211 designs sent for lab testing meeting the company's criteria for binding. Anthropic announced the release https://www.anthropic.com/research/claude-scorpion-toxin-binder-design?ref=runtimewire and published the designs and associated data https://huggingface.co/datasets/Anthropic/scorpion-toxin-binder-design?ref=runtimewire on October 20th, describing the work as an… When you purchase through links in our articles, we may earn a small commission https://runtimewire.com/affiliate-monetization-standards . This doesn’t affect our editorial independence.