EPFL Data Science Lab (dlab)

aiflows

🤖🌊 aiFlows: The building blocks of your collaborative AI

transformers-CFG

🤗 A specialized library for integrating context-free grammars (CFG) in EBNF with the Hugging Face Transformers

SynthIE

The data and the PyTorch implementation for the models and experiments in the paper "Exploiting Asymmetry for Synthetic Training Data Generation: SynthIE and the Case of Information Extraction".

homepage2vec

Language-Agnostic Website Embedding and Classification

cc_flows

The data and implementation for the experiments in the paper "Flows: Building Blocks of Reasoning and Collaborating AI".

GPTurk

llm-latent-language

Repo accompanying our paper "Do Llamas Work in English? On the Latent Language of Multilingual Transformers".

GCD

understanding-decoding

The data and the PyTorch implementation for the models and experiments in the paper "Language Model Decoding as Likelihood–Utility Alignment".

property-inference-attacks

Modular framework for property inference attacks on deep neural networks

forc

Framework for Cost-Effective Language Model Choice

LAMEN

transformers-GCD-PR

🤗 Transformers with Context-Free Grammar Generation Support

descartes

The PyTorch implementation for the models in the paper "Descartes: Generating Short Descriptions of Wikipedia Articles"

Negativity_in_2016_campaign

Code for the Paper "United States Politicians' Tone Became More Negative with 2016 Primary Campaigns"

distribution-inference-risks

Distribution Inference Risks: Identifying and Mitigating Sources of Leakage

AIReviewLottery

amplification_paradox

This repo contains the simulation code for the paper "The Amplification Paradox in Recommender Systems"

laughing-head

Code for the laughing head paper

llm-grounding-analysis

wiki_image_classification

Wikipedia Image Classification project

quotebank-toolkit

Scripts for cleaning and enriching Quotebank

deplatforming-parler

deplatforming_influencers

edisum

The data and the PyTorch implementation for the models and experiments in the paper "Edisum: Summarizing and Explaining Wikipedia Edits at Scale"

how_readers_browse_wikipedia

litellm-copy

Call all LLM APIs using the OpenAI format. Use Bedrock, Azure, OpenAI, Cohere, Anthropic, Ollama, Sagemaker, HuggingFace, Replicate (100+ LLMs)

multilingual-entity-insertion

Repository for "Fine-tuning large language models for link recommendation in Wikipedia" project

quotative_bias

youtube-embeddings

YouTube channel embeddings and social dimensions