Albert Thomas

Researcher, Huawei Technologies Ltd.

Joined

April 2020

Names

Albert Thomas (Preferred)

Emails

****@centraliens.net (Confirmed)

****@huawei.com (Confirmed)

Personal Links

Homepage

Google Scholar

DBLP

Career & Education History

Researcher

Huawei Technologies Ltd. (huawei.com)

2017 – Present

PhD student

Télécom ParisTech (telecom-paristech.fr)

2014 – 2017

Phd Student

Airbus (airbus.com)

2014 – 2017

Advisors, Relations & Conflicts

PhD Advisor

Stephan Clemencon

2014 – 2017

PhD Advisor

Alexandre Gramfort

2014 – 2017

Expertise

Model based Reinforcement learning

2020 – Present

Linear bandits

2018 – 2020

Best arm identification

2018 – 2020

Unsupervised anomaly detection

2014 – 2020

Publications

Embedding Distance as a Reward Signal can replace Verifiers for LLM Reasoning
Abdelhakim Benechehab, Youssef Attia El Hili, Albert Thomas, Giuseppe Paolo, Maurizio Filippone
- ICLR 2026 Workshop LLM Reasoning
- Readers: Everyone
From Data to Rewards: a Bi-level Optimization Perspective on Maximum Likelihood Estimation
Abdelhakim Benechehab, Gabriel Singer, Corentin Léger, Youssef Attia El Hili, Giuseppe Paolo, Albert Thomas, Maurizio Filippone, Balázs Kégl
- Submitted to ICLR 2026
- Readers: Everyone
LLMs as In-Context Meta-Learners for Model and Hyperparameter Selection
Youssef Attia El Hili, Albert Thomas, Malik Tiomoko, Abdelhakim Benechehab, Corentin Léger, Corinne Ancourt, Balázs Kégl
- Submitted to ICLR 2026
- Readers: Everyone
In-Context Meta-Learning with Large Language Models for Automated Model and Hyperparameter Selection
Youssef Attia El Hili, Albert Thomas, Abdelhakim Benechehab, Corentin Léger, Corinne Ancourt, Balázs Kégl
- NeurIPS 2025 LLM Evaluation Workshop Poster
- Readers: Everyone
AdaPTS: Adapting Univariate Foundation Models to Probabilistic Multivariate Time Series Forecasting
Abdelhakim Benechehab, Vasilii Feofanov, Giuseppe Paolo, Albert Thomas, Maurizio Filippone, Balázs Kégl
- SCOPE - ICLR 2025 Poster
- Readers: Everyone
AdaPTS: Adapting Univariate Foundation Models to Probabilistic Multivariate Time Series Forecasting
Abdelhakim Benechehab, Vasilii Feofanov, Giuseppe Paolo, Albert Thomas, Maurizio Filippone, Balázs Kégl
- ICML 2025 poster
- Readers: Everyone
Differentially Private Deep Model-Based Reinforcement Learning
Alexandre Rio, Merwan Barlier, Igor Colin, Albert Thomas
- Submitted to ICLR 2025
- Readers: Everyone
Zero-shot Model-based Reinforcement Learning using Large Language Models
Abdelhakim Benechehab, Youssef Attia El Hili, Ambroise Odonnat, Oussama Zekri, Albert Thomas, Giuseppe Paolo, Maurizio Filippone, Ievgen Redko, Balázs Kégl
- ICLR 2025 Poster
- Readers: Everyone
Differentially Private Deep Model-Based Reinforcement Learning
Alexandre Rio, Merwan Barlier, Igor Colin, Albert Thomas
- EWRL17
- Readers: Everyone
A Study of the Weighted Multi-step Loss Impact on the Predictive Error and the Return in MBRL
Abdelhakim Benechehab, Albert Thomas, Giuseppe Paolo, Maurizio Filippone, Balázs Kégl
- RLC 2024 ICBINB Poster
- Readers: Everyone

View all 24 publications

Co-Authors

View all 27 co-authors