Publications

DiffuMamba: High-Throughput Diffusion LMs with Mamba Backbone

Vaibhav Singh*, Oleksiy Ostapenko, Pierre-André Noël, Eugene Belilovsky, Torsten Scholak

International Conference on Machine Learning (ICML) (2026)

[Paper]

Dual-Phase Continual Learning: Supervised Adaptation Meets Unsupervised Retention

Vaibhav Singh*, Rahaf Aljundi, Eugene Belilovsky

Transactions on Machine Learning Research (TMLR) (2026)

[Paper]

Beyond Cosine Decay: On the Effectiveness of Infinite Learning Rate Schedule for Continual Pre-training

Vaibhav Singh*, Paul Janson*, Paria Mehrbod, Adam Ibrahim, Irina Rish, Eugene Belilovsky, Benjamin Therien

Conference on Lifelong Learning Agents (CoLLAs) (2025) Oral presentation

[Paper]

Model Parallelism With Subnetwork Data Parallelism

Vaibhav Singh*, Zafir Khalid, Edouard Oyallon, Eugene Belilovsky

ICML 2025 ESFOMO Workshop (2025)

[Paper]

Wake-Sleep Energy-Based Models for Continual Learning

Vaibhav Singh*, Anna Choromanska, Shuang Li, Yilun Du

CLVISION Workshop, CVPR 2024 (2024)

[Paper]

On Spectral and Temporal Feature Encoding Behaviour in Stacked Architectures

Vaibhav Singh*, Vinayak Abrol, Karan Nathwani

NeurIPS 2022 ENLSP Workshop (2022)

[Paper]