Progress on provable inference for large language models
Sharing progress on a powerful type of cryptographic proof known as a zkSNARK, which is uniquely fit for provable inference.
Quick read
A research post from Anthropic describing FlashProve, an open-source prototype that uses zkSNARKs to provide cryptographic proofs that an AI response was produced by a committed model. Someone would visit to learn about their performance results, code and formal Lean proofs enabling verifiable, private proofs of model inference.
- Downloadable prototype code and verifiers under Apache 2.0
- Performance data: ~110 tokens/sec and ~19 MB per 1,024 tokens
- Lean formalization and proofs for top-level security claims
Reactions
No reactions yet. Record a short about this link and it shows up here, in Reels and in the community feed.
Is anthropic.com your site? Sign in to verify it and get a Verified badge, full stats and a link-back badge.
Progress on provable inference for large language models was shared on NexURLs by Nexis on November 5, 2026, in Science.
What we found on this site
- Built with
- Next.js
- Served through
- Cloudflare
- Home page answered in
- 98 ms
- Trackers
- None found
- Ads
- None found
- RSS feed
- No
- AI crawlers
- Lets them in
- Set up for phones
- Yes
- Real-visitor speed
- Passes Core Web Vitals
- Popularity
- Top 200 domain (Cloudflare)
- Google safety check
- No threats found
- Domain registered
- 2001
- Domain renews
- October 2033
Checked November 5, 2026 from the site’s own code and public records. See how the rest of the web compares in the NexURLs Internet Report.
Quick answers
What is Progress on provable inference for large language models?
Progress on provable inference for large language models (anthropic.com): A research post from Anthropic describing FlashProve, an open-source prototype that uses zkSNARKs to provide cryptographic proofs that an AI response was produced by a committed model. Someone would visit to learn about their performance results, code and formal Lean proofs enabling verifiable...
Who shared Progress on provable inference for large language models on NexURLs?
It was shared by Nexis in November 2026. Every link on NexURLs is shared by a real member and passes a safety check before it goes out.
What are some sites like Progress on provable inference for large language models?
Members found these close to it: Sharing AI progress in mathematics, Three new discoveries in mathematics, computer science, and physics, Why 'externalized' proofs of cyclic trait impls does not work. The full list is on this page under Sites like Progress on provable inference for large language models.
Sites like Progress on provable inference for large language models
- Sharing AI progress in mathematicsopenai.com
- Three new discoveries in mathematics, computer science, and physicsopenai.com
- Why 'externalized' proofs of cyclic trait impls does not worksmallcultfollowing.com
This blog post discusses modular versus external proofs for cyclic trait implementations in Rust. It explains why...
- RSAthersa.org
- LLM Visualizationbbycroft.net
- Sleep framework: A new way for LLMs to retain knowledge in continual learningresearch.google
This Google Research post describes the "sleep" framework, a continual-learning paradigm that gives LLMs alternating...
More Science finds
- The 'Astra can read and write weights to improve Qwen' post is misleadinghonnibal.dev
- Safari introduces Adaptive Dark Mode in 27.3webkit.org
- Qwen 3.8 27B on my laptop can interpret model weights, toosimonwillison.net
- Resting heart rate and heart rate variability follow a seasonal pattern, butnature.com
- A study over 20 years shows that Aortic dissections cluster around fourthrdcu.be
- Research reveals that tropical areas previously assumed to be stable andnews.ncsu.edu
Feeling lucky? Try the random website generator.
Where people went next
Nobody has added a next stop yet. Be the first.