TEE Proof - Verify that AI providers can't read your prompts, even if they wanted to.
-
Updated
Nov 14, 2026 - Go
TEE Proof - Verify that AI providers can't read your prompts, even if they wanted to.
A framework for encrypted CKKS transformer inference on GPU. We implement complex slots utilization for decoding, packing awareness for sparse bootstrap placement, and a python API. Perseus supports encrypted language modelling, image and sentence classification, and image generation.
Example code that leverages Attested-TLS protocol (GENXT confido lib) to verify and access Confidential LLM API service.
CAPRI: Client-Assisted PRivate Inference. Run open-weight LLMs on an untrusted GPU server without it seeing your prompt or the answer. Lattice crypto only: one server, no trusted hardware.
PyTorch implementation of normalization-free LLMs investigating entropic behavior to find desirable activation functions
Research POC for masked private inference of public-weight LLMs without trusted hardware
A private DeFi trading agent using SolRouter encrypted inference - ensuring that your strategy queries never hit a public server in plaintext
To associate your repository with the private-inference topic, visit your repo's landing page and select "manage topics."