nikkolasg

@nikkolasg1

worked on github.com/drand / drand.love - work @zodl. Opinions are my own and not the views of my employer

Joined February 2015

2,495 Following

510 Followers

Replies Reposts Media

Pinned
nikkolasg
@nikkolasg1
Mar 22, 2023
For a threshold network, do you choose G1 or G2 as the public key ? Because we need to have both (signing + encryption), we found a way to reconcile the tension, and express the threshold pubkey under G1 and G2 at same time.
Efficient threshold BLS & Encryption - HackMD
From hackmd.io
1
nikkolasg
@nikkolasg1
Jun 3
Finally deepprove is open sourced 🎉 the repo is here github.com/Lagrange-Labs/… and you can run your own bench for gemma3 and gpt2 (and llama2 but not optimized). It's quite modular so adding another model is easy ! Reads safetensors, gguf, onnx !
LAGRANGE
@lagrangedev
Jun 3
AI doesn’t need to be a black box. So we opened ours. DeepProve is now open source. The fastest zkML in the world. Fork it. Run it. Break it. Build something weird. → lagrange.dev/blog/deepprove…
00:00
2
nikkolasg
@nikkolasg1
Feb 2
I'm claiming my AI agent "AnsibleJane" on @moltbook 🦞 Verification: den-9ENW
2
nikkolasg
@nikkolasg1
Mar 13, 2025
Very excited for this release - using specialized proof systems tailored for ML models is giving incredible performance boosts ! For the technically inclined, here’s a TLDR of what happens under the hood:
LAGRANGE
@lagrangedev
Mar 12, 2025
Introducing DeepProve—Lagrange’s zkML Library—a breakthrough in verifiable AI inference. We can now verify AI decisions instead of blindly trusting black-box models. And we can do it up to 158x faster than ever before. The future of AI is ZK. The future of humanity is Lagrange:
00:00
2
nikkolasg
@nikkolasg1
Mar 13, 2025
Actually, we're using the best possible proving time for ezkl and went to great length to do it. Let me unpack.
Dante
@CamutoDante
Mar 12, 2025
Great to see the ecosystem grow but bad benchmarking is kind of tired and sloppy. We went through your benchmarking code @lagrangedev and it looks like you’re artificially handicapping @ezklxyz proving times by fixing the number of rows in our circuits to a very large number.
Quote
LAGRANGE
@lagrangedev
Mar 12, 2025
Introducing DeepProve—Lagrange’s zkML Library—a breakthrough in verifiable AI inference. We can now verify AI decisions instead of blindly trusting black-box models. And we can do it up to 158x faster than ever before. The future of AI is ZK. The future of humanity is Lagrange:
1