Anes Benmerzoug
Anes is an AI Engineer with background in Control Engineering and Embedded Systems. He first dipped his feet in the Machine Learning field at the end of his second Master’s thesis. After that he went on to work in the field for an antivirus company and finally ended up at appliedAI in 2021. Since then, he has worked on many projects and is always looking forward to interesting things to work on.

Anes' work

CLP-Transfer: Cross-Lingual and Progressive Transfer Learning

The CLP-Transfer method introduces a novel approach for cross-lingual language transfer by leveraging token overlap and a small pre-trained …

Anes Benmerzoug
Large Language Models
Pill

WECHSEL: Cross-Lingual Transfer

Language transfer enables the use of language models trained in one or more languages to initialize a new language model in another …

Anes Benmerzoug
Large Language Models
Pill

Variance Reduced Shapley Value Estimation for Trustworthy Data Valuation

VRDS, an effective method to reduce the variance of Data Shapley, based on a stratified sampling technique with precomputed coefficients.

Anes Benmerzoug
Data Valuation
Training

Safe and efficient deep reinforcement learning

This 2-day dive into deep RL for decision and control is suitable for engineers who want to solve real-world control problems with efficient …

Training

Machine Learning and Control

A one-day training introducing concepts and principles from control theory, with a heavy focus on optimal control theory, and presenting its …

Seminar

Shapley values for XAI: the good, the bad and the ugly

In this talk Anes will ask questions like: What is the true significance of Shapley values as feature importance measures? How can Shapley …

Anes Benmerzoug
Explainable AI
Pill

Measuring the Intrinsic Dimension of Objective Landscapes

This paper proposes a method for measuring the difficulty of a machine learning task by training neural networks using a randomly oriented …

Anes Benmerzoug
Learning in High Dimensions
Software

pyDVL: the python Data Valuation Library

pyDVL strives to offer reference implementations of algorithms for data valuation and influence function computation, with an emphasis on …

Anes Benmerzoug Miguel de Benito Delgado Janoś Gabler Fabio Peruzzo Kristof Schröder Markus Semmler
Data Valuation
Training

Methods and issues in explainable AI

A two-day workshop for ML practitioners wishing to make their models better understandable for themselves and decision makers.