VerInf github.com ↗
An open-source prototype for proving LLM inference, including Llama 4 Maverick.
LibrarySecurity & privacyzero-knowledgellm-inferenceverificationzero-knowledge-proofs
Captured
Recommended in 1 thread
Progress on provable inference for large language modelsanthropic.com ·
Its author says the Anthropic work builds directly on it.
Similar tools
- pledge ↗Security & privacy0 threads
An OpenBSD system call that irrevocably restricts the operations a process can perform.
- 2 threads
- AppArmor ↗Security & privacy1 thread
A Linux kernel security module that confines programs with per-application profiles.
- 1 thread
- 1 thread
- matchlock ↗Security & privacy1 thread
A Firecracker-based VM sandbox for agents that injects secrets into network requests via placeholders.
- 1 thread
- 2 threads
Also in Security & privacy
- 1Password ↗8 threads
A commercial password manager with a built-in SSH agent that can be unlocked with Touch ID.
- 5 threads
- 4 threads
- 3 threads
- Signal3 threads
An end-to-end encrypted messaging app.
- Tor3 threads
An anonymity network that routes traffic through volunteer relays.
- 2 threads
- 2 threads
- 2 threads
- 2 threads
- Convergence ↗2 threads
Moxie Marlinspike's 2011 proposal for client-chosen notary-based TLS certificate verification.
- Darknet Diaries ↗2 threads
A podcast covering true stories from the dark side of the internet, including hacking and cyber heists.
Link from a commenter, not checked or endorsed by us: ↗ https://github.com/jamespetrie/verinf