Gopeshh Subbaraj

Matthew D Riemer

Nizar Islah

Benjamin Therien

Tsuguchika Tabaru

Hiroaki Kingetsu

Sarath Chandar

Training large language models (LLMs) typically involves pre-training on massive corpora, only to restart the process entirely when new data… (see more) becomes available. A more efficient and resource-conserving approach would be continual pre-training, where models are updated with new data rather than retraining from scratch. However, the introduction of new data often causes distribution shifts, leading to performance degradation on previously learned tasks. In this paper, we take a deeper look at two popular proposals for addressing this distribution shift within the continual learning literature: experience replay and gradient alignment. We consider continual pre-training of models within the Llama family of architectures at a large scale across languages with 100 billion tokens of training data in each language, finding that both replay and gradient alignment lead to more stable learning without forgetting. This conclusion holds both as we vary the model scale and as we vary the number and diversity of tasks. Moreover, we are the first to demonstrate the effectiveness of gradient alignment techniques in the context of LLM pre-training and propose an efficient implementation of meta-experience replay (MER) that imbues experience replay with the benefits of gradient alignment despite negligible compute and memory overhead. Our scaling analysis across model sizes and replay rates indicates that small rates of replaying old examples are definitely a more valuable use of compute than investing in model size, but that it is more compute efficient to scale the size of the model than invest in high rates of replaying old examples.

2025-08-03

ArXiv (preprint)

Revisiting Replay and Gradient Alignment for Continual Pre-Training of Large Language Models

Istabrak Abbes

Matthew D Riemer

Nizar Islah

Benjamin Therien

Tsuguchika Tabaru

Hiroaki Kingetsu

Sarath Chandar

2025-08-03

ArXiv (preprint)

Behavioral Suite Analysis of Self-Supervised Learning in Atari

Somjit Nath

Rishav

Derek Nowrouzezahrai

Samira Ebrahimi Kahou

2025-06-20

rl-conference.cc/RLC/2025/Workshop/RLVG (accepted)

Pretraining Generative Flow Networks with Inexpensive Rewards for Molecular Graph Generation

Mohit Pandey

Artem Cherkasov

Martin Ester

Emmanuel Bengio

2025-03-01

arXiv (published)

Enabling Realtime Reinforcement Learning at Scale with Staggered Asynchronous Inference

Matthew D Riemer

Glen Berseth

Realtime environments change even as agents perform action inference and learning, thus requiring high interaction frequencies to effectivel… (see more)y minimize regret. However, recent advances in machine learning involve larger neural networks with longer inference times, raising questions about their applicability in realtime systems where reaction time is crucial. We present an analysis of lower bounds on regret in realtime reinforcement learning (RL) environments to show that minimizing long-term regret is generally impossible within the typical sequential interaction and learning paradigm, but often becomes possible when sufficient asynchronous compute is available. We propose novel algorithms for staggering asynchronous inference processes to ensure that actions are taken at consistent time intervals, and demonstrate that use of models with high action inference times is only constrained by the environment's effective stochasticity over the inference horizon, and not by action frequency. Our analysis shows that the number of inference processes needed scales linearly with increasing inference times while enabling use of models that are multiple orders of magnitude larger than existing approaches when learning from a realtime simulation of Game Boy games such as Pokémon and Tetris.

2025-01-22

ICLR.cc/2025/Conference (poster)

Seq-VCR: Preventing Collapse in Intermediate Transformer Representations for Enhanced Reasoning

Md Rifat Arefin

Nicolas Gontier

Yann LeCun

Ravid Shwartz-Ziv

Chris Pal

2025-01-22

ICLR.cc/2025/Conference (poster)

GFlowNet Pretraining with Inexpensive Rewards

Mohit Pandey

Emmanuel Bengio

Generative Flow Networks (GFlowNets), a class of generative models have recently emerged as a suitable framework for generating diverse and … (see more)high-quality molecular structures by learning from unnormalized reward distributions. Previous works in this direction often restrict exploration by using predefined molecular fragments as building blocks, limiting the chemical space that can be accessed. In this work, we introduce Atomic GFlowNets (A-GFNs), a foundational generative model leveraging individual atoms as building blocks to explore drug-like chemical space more comprehensively. We propose an unsupervised pre-training approach using offline drug-like molecule datasets, which conditions A-GFNs on inexpensive yet informative molecular descriptors such as drug-likeliness, topological polar surface area, and synthetic accessibility scores. These properties serve as proxy rewards, guiding A-GFNs towards regions of chemical space that exhibit desirable pharmacological properties. We further our method by implementing a goal-conditioned fine-tuning process, which adapts A-GFNs to optimize for specific target properties. In this work, we pretrain A-GFN on the ZINC15 offline dataset and employ robust evaluation metrics to show the effectiveness of our approach when compared to other relevant baseline methods in drug design.

2024-09-15

ArXiv (preprint)

GFlowNet Pretraining with Inexpensive Rewards

Mohit Pandey

Emmanuel Bengio

Generative Flow Networks (GFlowNets), a class of generative models have recently emerged as a suitable framework for generating diverse and … (see more)high-quality molecular structures by learning from unnormalized reward distributions. Previous works in this direction often restrict exploration by using predefined molecular fragments as building blocks, limiting the chemical space that can be accessed. In this work, we introduce Atomic GFlowNets (A-GFNs), a foundational generative model leveraging individual atoms as building blocks to explore drug-like chemical space more comprehensively. We propose an unsupervised pre-training approach using offline drug-like molecule datasets, which conditions A-GFNs on inexpensive yet informative molecular descriptors such as drug-likeliness, topological polar surface area, and synthetic accessibility scores. These properties serve as proxy rewards, guiding A-GFNs towards regions of chemical space that exhibit desirable pharmacological properties. We further our method by implementing a goal-conditioned fine-tuning process, which adapts A-GFNs to optimize for specific target properties. In this work, we pretrain A-GFN on the ZINC15 offline dataset and employ robust evaluation metrics to show the effectiveness of our approach when compared to other relevant baseline methods in drug design.

2024-09-15

ArXiv (preprint)

Realtime Reinforcement Learning: Towards Rapid Asynchronous Deployment of Large Models

Matthew D Riemer

Glen Berseth

Realtime environments change even as agents perform action inference and learning, thus requiring high interaction frequencies to effectivel… (see more)y minimize long-term regret. However, recent advances in machine learning involve larger neural networks with longer inference times, raising questions about their applicability in realtime systems where reaction time is crucial. We present an analysis of lower bounds on regret in realtime environments to show that minimizing long-term regret is generally impossible within the typical sequential interaction and learning paradigm, but often becomes possible when sufficient asynchronous compute is available. We propose novel algorithms for staggering asynchronous inference processes to ensure that actions are taken at consistent time intervals, and demonstrate that use of models with high action inference times is only constrained by the environment's effective stochasticity over the inference horizon, and not by action frequency. Our analysis shows that the number of inference processes needed scales linearly with increasing inference times while enabling use of models that are multiple orders of magnitude larger than existing approaches when learning from a realtime simulation of Game Boy games such as Pokemon and Tetris.

2024-06-19

ICML.cc/2024/Workshop/ARLET (poster)

Discovering Object-Centric Generalized Value Functions From Pixels

Somjit Nath

Khimya Khetarpal

Samira Ebrahimi Kahou

Deep Reinforcement Learning has shown significant progress in extracting useful representations from high-dimensional inputs albeit using ha… (see more)nd-crafted auxiliary tasks and pseudo rewards. Automatically learning such representations in an object-centric manner geared towards control and fast adaptation remains an open research problem. In this paper, we introduce a method that tries to discover meaningful features from objects, translating them to temporally coherent"question"functions and leveraging the subsequent learned general value functions for control. We compare our approach with state-of-the-art techniques alongside other ablations and show competitive performance in both stationary and non-stationary settings. Finally, we also investigate the discovered general value functions and through qualitative analysis show that the learned representations are not only interpretable but also, centered around objects that are invariant to changes across tasks facilitating fast adaptation.

2023-07-03

Proceedings of the 40th International Conference on Machine Learning (published)

Continual Learning In Environments With Polynomial Mixing Times

Matthew D Riemer

Sharath Chandra Raparthy

Ignacio Cases

Maximilian Puelma Touzel

The mixing time of the Markov chain induced by a policy limits performance in real-world continual learning scenarios. Yet, the effect of mi… (see more)xing times on learning in continual reinforcement learning (RL) remains underexplored. In this paper, we characterize problems that are of long-term interest to the development of continual RL, which we call scalable MDPs, through the lens of mixing times. In particular, we theoretically establish that scalable MDPs have mixing times that scale polynomially with the size of the problem. We go on to demonstrate that polynomial mixing times present significant difficulties for existing approaches that suffer from myopic bias and stale bootstrapped estimates. To validate the proposed theory, we study the empirical scaling behavior of mixing times with respect to the number of tasks and task switching frequency for pretrained high performing policies on seven Atari games. Our analysis demonstrates both that polynomial mixing times do emerge in practice and how their existence may lead to unstable learning behavior like catastrophic forgetting in continual learning settings.