
DeepProve proves which AI model answered you. Cited by @AnthropicAI. Halo keeps robots and drones in sync when GPS and radio go down. @LagrangeFndn $LA
Los Angeles
Joined May 2022
- Tweets3.9K
- Following387
- Followers93.5K
- Likes2.7K
Pinned Tweet
The Frontier Red Team at @AnthropicAI just made the case that you should be able to check which AI model answered you, without ever seeing its weights. Their research names DeepProve “the current peer-reviewed state of the art.”
The US banned @deepseek_ai from government software in an order signed in July and published in October. The deadline to ask for more time passed a month before anyone could read it. If a vendor had DeepSeek inside its software, how would you know?
It reaches other companies' software too, and nobody has said how to check. lagrange.dev/blog/the-us-ban…
2
183
For anyone not reading the paper before Wednesday: it proves every token of a model’s full answer, and checking the proof takes 1 to 3.7 seconds.
Replying to @chbpap
@chbpap presents our deepprove paper at ACM CCS in the Hague on wednesday, the one @AnthropicAI called the current peer-reviewed state of the art.
If you’re there, go find him. what do you want to ask him?
AI made images, video and text in Europe now have to carry a label saying so. The label gets added by whoever made them. It can't prove which AI made them, or that it's true. Would a label change whether you believe what you're seeing?
LAGRANGE retweeted
ai seemingly doesn’t know which model it is, but it’s very sure about literally everything else.
A player jammed the swarm with Halo 35 times in under a minute. Three times it worked out where the jamming came from and flew around it.
Tell us the best places you'd hide a jammer?
holdtheswarm.lagrange.dev
Sriram's paper proves the curved math inside machine learning more accurately, without the extra cost that usually comes with it. Under two minutes. Ask us anything about it below.
The paper, by Sriram Sridhar, Shravan Srinivasan, Dimitrios Papadopoulos and @chbpap: eprint.iacr.org/2025/2326
366
It's fixable: a proof that comes with every answer shows exactly which model produced it.
Which AI tool at work would you want to check first?
Last week, we asked four AI apps which model they are. Right now you just have to take their word for it. Ask yours and post what it says.
If a routing service quietly swapped the model answering you for a cheaper one, nothing in the reply would tell you. Anthropic's red team used that exact example last week to make the case for ZK.
What @AnthropicAI built and where DeepProve fits ⬇️
I've seen @lagrangedev and StarkNet being mentioned here, but only a few know the connection between this article by @AnthropicAI and Rari Capital, if anyone here still remembers that.
Significant motion on part of 2020/2021 DeFi devs, a cohort amazingly abundant in s++ talent
teapawt.org/AnthropicAI/status/211…
New Anthropic research: progress on provable inference for large language models.
Here, we introduce FlashProve. Building on established cryptographic techniques, FlashProve allows developers to verify whether a given response came from a given AI model. It’s substantially more efficient than prior open-source methods for doing so.
These proofs have a range of applications. They can, for instance, make it harder for attackers to hijack a chat, or enable safety evaluators to verify that the model they’re testing is the same one being served to customers.
For our full write-up, see here: anthropic.com/research/prova…
Further progress in provable inference will likely be an important part of improving trust in AI deployment. To that end, we're open-sourcing FlashProve’s code, here: github.com/anthropics/FlashP…
We’re grateful to our collaborators: Alessandro Chiesa, @GiacomoFenzi, @gwberg95, @miranda__christ, Rahul Dalal (rdalal.github.io) and Ben Mayer (@transmissions11).
LAGRANGE retweeted
클로드(Claude) 개발사 앤트로픽(Anthropic)에서
라그랑주 딥프로브(DeepProve)를 언급했습니다!!!
@AnthropicAI가 공개한 AI 추론 검증 연구에
@lagrangedev의 DeepProve가 소개됐습니다.
선행 연구를 다루는 부분에서
'동료 심사를 거친 현재 최고 수준의 기술 중 하나'라고
평가한 점이 눈에 들어오는데요.
어떤 기술인지 쉽게 풀어보면 이렇습니다.
"지금 받은 답변, 정말 내가 선택한 AI 모델이 만든 게 맞을까?"
AI 서비스를 이용할 때
실제로 어떤 모델이 답변했는지
사용자가 직접 확인하기는 어렵습니다.
딥프로브(DeepProve)는
바로 이 부분을 다룹니다.
답변 전체가 지정된 모델의 계산 결과인지
영지식증명으로 검증하는 기술입니다.
앤트로픽(Anthropic)은 이번 연구에서
딥프로브(DeepProve)의 검증 방식과
처리 성능도 함께 소개했습니다.
AI가 점점 더 많은 일을 맡게 되는 만큼,
어떤 모델이 그 결과를 만들었는지
확인하는 기술도 중요해지는 것 같습니다.
라그랑주 앰배서더로서 정말 반가운 소식이네요.
꾸준히 쌓아온 연구가 이런 곳에서 언급되는 모습을 보니
앞으로의 발전도 더 기대하게 됩니다.
🔗 앤트로픽(Anthropic) 공식 연구 : anthropic.com/research/prova…)
🔗 딥프로브(DeepProve) 논문 : eprint.iacr.org/2026/1112
🤖 Made with AI
Proud that @AnthropicAI’s research recognizes @lagrangedev's DeepProve as the current peer-reviewed state of the art.
Every AI answer should come with a proof of which model gave it. Trust is moving from “take our word for it” to “check it yourself,” and today a frontier lab started agreeing.
A player jammed our swarm 389 times over Los Angeles, for more than ten minutes. It lost no drones and finished all eight sites 30 seconds after the jamming stopped. How long can you hold it?
holdtheswarm.lagrange.dev
For most of computing, checking someone's work meant doing it again.
You cannot do that with an AI inference. You do not have the weights. You do not have the input data. You do not have their hardware. Re-running is not expensive, it is unavailable.
That is why a proof is not a convenience. It is the only check left.
Anthropic's Frontier Red Team made the same argument this week.
No.3, who made the record.
teapawt.org/lagrangedev/status/210…
Every trust mechanism in software comes down to one question. Who made the record.
A log is made by the system, afterwards, and it says what the system decided to write. An audit is made by a person, later, out of those logs. A proof is made by the computation, while it runs.
Nobody writes it. Nobody can edit it. That is the whole difference.
Prove you're over 18 without showing your birthday. Prove the answer is right without showing the work. DeepProve does that for AI inference, and @AnthropicAI's research team just named it the current peer-reviewed state of the art in the field. @chbpap explains the idea.
A model gets retrained on a Tuesday and every proof about it has to start over. @chbpap, our chief scientist, walks through the Eurocrypt 2026 paper on proofs that update instead.
Proud that @AnthropicAI’s research recognizes @lagrangedev's DeepProve as the current peer-reviewed state of the art.
Every AI answer should come with a proof of which model gave it. Trust is moving from “take our word for it” to “check it yourself,” and today a frontier lab started agreeing.
New Anthropic research: progress on provable inference for large language models.
Here, we introduce FlashProve. Building on established cryptographic techniques, FlashProve allows developers to verify whether a given response came from a given AI model. It’s substantially more efficient than prior open-source methods for doing so.
These proofs have a range of applications. They can, for instance, make it harder for attackers to hijack a chat, or enable safety evaluators to verify that the model they’re testing is the same one being served to customers.
For our full write-up, see here: anthropic.com/research/prova…
Further progress in provable inference will likely be an important part of improving trust in AI deployment. To that end, we're open-sourcing FlashProve’s code, here: github.com/anthropics/FlashP…
We’re grateful to our collaborators: Alessandro Chiesa, @GiacomoFenzi, @gwberg95, @miranda__christ, Rahul Dalal (rdalal.github.io) and Ben Mayer (@transmissions11).
Big congrats to @chbpap and the team behind DeepProve. Read the paper: eprint.iacr.org/2026/1112
5
444











