ServiceNow IA recherche

Reinforcement Learning

Destruction is a General Strategy to Learn Generation; Diffusion's Strength is to Take it Seriously; Exploration is the Future
Destruction is a General Strategy to Learn Generation; Diffusion's Strength is to Take it Seriously; Exploration is the Future
I present diffusion models as part of a family of machine learning techniques that withhold information from a model’s input and train …
ICLR Blogposts 2026,  2026.
Article Citation
PipelineRL: Faster On-policy Reinforcement Learning for Long Sequence Generation
PipelineRL: Faster On-policy Reinforcement Learning for Long Sequence Generation
Transactions on Machine Learning Research (TMLR),  2026.
Article Citation Code
LLMs can learn self-restraint through iterative self-reflection
LLMs can learn self-restraint through iterative self-reflection
In order to be deployed safely, Large Language Models (LLMs) must be capable of dynamically adapting their behavior based on their …
Transactions on Machine Learning Research (TMLR),  2025.
Article Citation Vidéo
Multimodal foundation world models for generalist embodied agents
Multimodal foundation world models for generalist embodied agents
Learning generalist agents, able to solve multitudes of tasks in different domains is a long-standing problem. Reinforcement learning …
Neural Information Processing Systems (NeurIPS),  2024.
Article Citation Code
Representing Positional Information in Generative World Models for Object Manipulation
Representing Positional Information in Generative World Models for Object Manipulation
The ability to predict outcomes of interactions between embodied agents and objects is paramount in the robotic setting. While …
Workshop at the Neural Information Processing Systems (NeurIPS),  2024.
Article Citation
Self-evaluation and self-prompting to improve the reliability of LLMs
Self-evaluation and self-prompting to improve the reliability of LLMs
In order to safely deploy Large Language Models (LLMs), they must be capable of dynamically adapting their behavior based on their …
Workshop at the International Conference of Learning Representation (ICLR),  2024.
Article Citation Vidéo
Bridging the Gap Between Target Networks and Functional Regularization
Bridging the Gap Between Target Networks and Functional Regularization
Target networks are at the core of recent success in Reinforcement Learning. They stabilize the training by using old parameters to …
Transactions on Machine Learning Research (TMLR),  2023.
Article Citation Code
Using Confounded Data in Latent Model-Based Reinforcement Learning
Using Confounded Data in Latent Model-Based Reinforcement Learning
In the presence of confounding, naively using off-the-shelf offline reinforcement learning (RL) algorithms leads to sub-optimal …
Transactions on Machine Learning Research (TMLR),  2023.
Article Citation Code
Mastering the Unsupervised Reinforcement Learning Benchmark from Pixels
Mastering the Unsupervised Reinforcement Learning Benchmark from Pixels
Controlling artificial agents from visual sensory data is an arduous task. Reinforcement learning (RL) algorithms can succeed but …
International Conference on Machine Learning (ICML),  2023.
Article Citation Code
Choreographer: Learning and Adapting Skills in Imagination
Choreographer: Learning and Adapting Skills in Imagination
Unsupervised skill learning aims to learn a rich repertoire of behaviors without external supervision, providing artificial agents with …
International Conference of Learning Representations (ICLR),  2023.
Article Citation Code