Stars
Using LLMs for efficient communication of explainers.
Preswald is a WASM packager for Python-based interactive data apps: bundle full complex data workflows, particularly visualizations, into single files, runnable completely in-browser, using Pyodide…
Explainable clustering method used to extract interpretable student profiles. Featured at AIED 2024.
InterpretCC, an inherently interpretable neural network architecture
Meditron is a suite of open-source medical Large Language Models (LLMs).
MultiModN – Multimodal, Multi-Task, Interpretable Modular Networks (NeurIPS 2023)
Landmark Attention: Random-Access Infinite Context Length for Transformers
Validating post-hoc explainers with learning science experts (LAK 2023).
Interpretability on raw time series with graph neural nets and concept activation vectors. Featured at AAAI 2023.
Bias analysis of german language models (BERT, T5, GPT2), with a case study on educational peer review data.
Success prediction in flipped classrooms via clickstreams. Published at EDM 2021.
Code for the paper "Evolutionary Clustering of Apprentices' Behavior in Online Learning Journals for Vocational Education" published at SFUVET 2022.
Comparing 5 different XAI techniques (LIME, PermSHAP, KernelSHAP, DiCE, CEM) through quantitative metrics. Published at EDM 2022.
MIDI2vec computes embeddings for representing MIDI data in vector space
A collection of pre-trained, state-of-the-art models in the ONNX format
[CVPR 2020] Normal Assisted Stereo Depth Estimation
Models to warm-start student pass/fail performance prediction trained using meta-transfer learning on 26 courses, 100k students, and millions of interactions. Published at L@S 2022.
Lab Materials for the EPFL Course on Machine Learning for Behavioral Data CS-421, Spring 2022.
🤖 Machine Learning Summer School Guide
Guidebook for IC PhD life at EPFL
Python notebooks with ML and deep learning examples with Azure Machine Learning Python SDK | Microsoft
Identify which instances in a database represent the same real-world entity (entity resolution). SIGMOD 2021.
Style Transfer for non-parallel text
Extracting knowledge graphs from language models as a diagnostic benchmark of model performance (NeurIPS XAI 2021).
This repository contains implementations and illustrative code to accompany DeepMind publications
Practical low-rank gradient compression for distributed optimization: https://arxiv.org/abs/1905.13727