SRI Lab, ETH Zurich

lmql

A language for constraint-guided and efficient LLM programming.

silq

debin

Machine Learning to Deobfuscate Binaries

diffai

A certifiable defense against adversarial examples by training neural networks to be provably robust

language-model-arithmetic

Controlled Text Generation via Language Model Arithmetic

ilf

AI based fuzzer based on imitation learning

psi

Exact Inference Engine for Probabilistic Programs

sven

dl2

DL2 is a framework that allows training neural networks with logical constraints over numerical values in the network (e.g. inputs, outputs, weights) and to query networks for inputs fulfilling a logical formula.

fastsmt

Learning to Solve SMT Formulas Fast

llmprivacy

ChatProtect

This is the code for the paper "Self-contradictory Hallucinations of Large Language Models: Evaluation, Detection and Mitigation".

SafeCoder

SynthPAI

A Synthetic Dataset for Personal Attribute Inference

watermark-stealing

Watermark Stealing in Large Language Models (ICML '24)

tableak

TabLeak: Tabular Data Leakage in Federated Learning

llm-quantization-attack

eth-sri.github.io

SRI Group Website

mn-bab

[ICLR 2022] Complete Verification via Multi-Neuron Relaxation Guided Branch-and-Bound

cuts

privacy-inference-multimodal

fairness-feedback-nlp

Human-Guided Fair Classification for NLP (ICLR 2023, Spotlight)

ChromeER

ConStat

A statistical test for contamination detection in language models.

datamin

This is the Repository containing the implementation of our paper "From Principle to Practice: Vertical Data Minimization for Machine Learning".

synthetiq

domino

watermark-stealing-website

The website for "Watermark Stealing in Large Language Models".

constat.github.io

CTBench