Publications

Beyond Autoregression: Fast LLMs via Self-Distillation Through Time

J. S. Deschenaux; C. Gulcehre 

Proceedings of the Thirteenth International Conference on Learning Representations (ICLR) 2025 [Forthcoming publication]

2025

13th International Conference on Learning Representations (ICLR 2025), Singapore, 2025-04-24 – 2025-04-28.

Building on Efficient Foundations: Effectively Training LLMs with Structured Feedforward Layers

X. Wei; S. Moalla; R. Pascanu; C. Gulcehre 

Advances in Neural Information Processing Systems 37 (NeurIPS 2024)

2024

38th Annual Conference on Neural Information Processing Systems, Vancouver Convention Center, 2024-12-10 – 2024-12-15.

The Effect of Scheduling and Preemption on the Efficiency of LLM Inference Serving

K-M. Kim; K. Hong; C. Gulcehre; A. Ailamaki 

2024

p. 16.

No Representation, No Trust: Connecting Representation, Collapse, and Trust Issues in PPO

S. Moalla; R. Pascanu; A. S. A. Miele; D. Pyatko; C. Gulcehre 

Advances in Neural Information Processing Systems 37 (NeurIPS 2024)

2024

38th Annual Conference on Neural Information Processing Systems, Vancouver Convention Center, 2024-12-10 – 2024-12-15.

Self-Recognition in Language Models

T. R. Davidson; V. Surkov; V. Veselovskyy; G. Russo; R. West et al. 

Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing

2024

Conference on Empirical Methods in Natural Language Processing (EMNLP 2024), Miami, Florida, USA, 2024-11-12 – 2024-11-16.

p. 12032 – 12059

SIMPLE HIERARCHICAL PLANNING WITH DIFFUSION

C. Chen; D. Fei; K. Kenji; C. Gulcehre; A. Sungjin 

2024

ICLR 2024, Vienna, Austria, May 7th to 11th, 2024.

Imagine the Unseen World: A Benchmark for Systematic Generalization in Visual World Models

Y. Kim; G. Singh; J. Park; C. Gulcehre; S. Ahn 

37th Conference on Neural Information Processing Systems (NeurIPS 2023) Track on Datasets and Benchmarks

2023

37th Annual Conference on Neural Information Processing Systems, New Orleans, USA, 2023-12-10 – 2023-12-16.

DOI : 10.48550/arXiv.2311.09064