Goto

Collaborating Authors

 Deep Learning

2b47305e1c81890b1089a405686ad183-Paper-Conference.pdf

QA: A Question-Answering Dataset for Benchmarking LLMs on Unseen Reference Content João Monteiro,3, Pierre-André Noël

Neural Information Processing Systems

Consequently, evaluating models on test splits that might have leaked into the training set is prone to misleading conclusions.

2b0e14abd8128e6bf98b6b0bec1cfcbf-Paper-Conference.pdf

MiSO: Optimizing brain stimulation to create neural population activity states

Neural Information Processing Systems

Brain stimulation has the potential to create desired neural population activity states.

Diffusion Forcing: Next-token Prediction Meets Full-Sequence Diffusion

Neural Information Processing Systems

Work done as a visiting student at MIT. 38th Conference on Neural Information Processing Systems (NeurIPS 2024).

2ae6b2bdf3a179e3e24129e2c54bd871-Paper-Conference.pdf

Be like a Goldfish, Don't Memorize! Mitigating Memorization in Generative LLMs

The Implicit Bias of Adam on Separable Data

Neural Information Processing Systems

Adam has become one of the most favored optimizers in deep learning problems.

Vision Model Pre-training on Interleaved Image-Text Data via Latent Compression Learning Chenyu Y ang

Neural Information Processing Systems

Raw pixels are unstructured and often contain unnecessary and unpredictable details.

If You Want to Be Robust, Be Wary of Initialization

Neural Information Processing Systems

We introduce a theoretical framework bridging the connection between initialization strategies and a network's resilience to adversarial perturbations.