AIConfPaper
← Search

Benjamin Negrevergne

8 accepted papers

2025

Improving Diversity in Language Models: When Temperature Fails, Change the Loss

Alexandre Verine, Florian Le Bronnec, Kunhao Zheng, Alexandre Allauzen, Yann Chevaleyre, benjamin negrevergne
ICML 2025poster

Increasing diversity in language models is a challenging yet essential objective. A common approach is to raise the decoding temperature. In this work, we investigate this approach through a simplistic yet common case to provide insights into why decreasing temperature can improve quality (Precision…

Cited by 0Source ↗PDF ↗Scholar ↗
2025

Unveiling the Role of Randomization in Multiclass Adversarial Classification: Insights from Graph Theory

Lucas Gnecco Heredia, Matteo Sammut, Muni Sreenivas Pydi, Rafael Pinot, benjamin negrevergne, Yann Chevaleyre
AISTATS 2025poster

Randomization as a mean to improve the adversarial robustness of machine learning models has recently attracted significant attention. Unfortunately, much of the theoretical analysis so far has focused on binary classification, providing only limited insights into the more complex multiclass setting…

Cited by 0Source ↗Scholar ↗
2025

What Makes Large Language Models Reason in (Multi-Turn) Code Generation?

Kunhao Zheng, Juliette Decugis, Jonas Gehring, Taco Cohen, benjamin negrevergne, Gabriel Synnaeve
ICLR 2025poster

Prompting techniques such as chain-of-thought have established themselves as a popular vehicle for improving the outputs of large language models (LLMs). For code generation, however, their exact mechanics and efficacy are under-explored using unified metrics and benchmarks. We thus investigate the…

Cited by 7Source ↗PDF ↗Scholar ↗
2024

Exploring Precision and Recall to assess the quality and diversity of LLMs

Florian Le Bronnec, Alexandre Verine, Benjamin Negrevergne, Yann Chevaleyre, Alexandre Allauzen
ACL 2024long

We introduce a novel evaluation framework for Large Language Models (LLMs) such as Llama-2 and Mistral, focusing on importing Precision and Recall metrics from image generation to text generation. This approach allows for a nuanced assessment of the quality and diversity of generated text without th…

Cited by 15Source ↗PDF ↗code ↗Scholar ↗
2024

Optimal Budgeted Rejection Sampling for Generative Models

Alexandre Verine, Muni Sreenivas Pydi, Benjamin Negrevergne, Yann Chevaleyre
AISTATS 2024poster

Rejection sampling methods have recently been proposed to improve the performance of discriminator-based generative models. However, these methods are only optimal under an unlimited sampling budget, and are usually applied to a generator trained independently of the rejection procedure. We first pr…

Cited by 5Source ↗PDF ↗Scholar ↗
2023

On the Role of Randomization in Adversarially Robust Classification

Lucas Gnecco Heredia, Muni Sreenivas Pydi, Laurent Meunier, benjamin negrevergne, Yann Chevaleyre
NeurIPS 2023spotlight

Deep neural networks are known to be vulnerable to small adversarial perturbations in test data. To defend against adversarial attacks, probabilistic classifiers have been proposed as an alternative to deterministic ones. However, literature has conflicting findings on the effectiveness of probabili…

Cited by 5Source ↗PDF ↗Scholar ↗
2023

Precision-Recall Divergence Optimization for Generative Modeling with GANs and Normalizing Flows

Alexandre Verine, benjamin negrevergne, Muni Sreenivas Pydi, Yann Chevaleyre
NeurIPS 2023poster

Achieving a balance between image quality (precision) and diversity (recall) is a significant challenge in the domain of generative models. Current state-of-the-art models primarily rely on optimizing heuristics, such as the Fr\'echet Inception Distance. While recent developments have introduced pri…

Cited by 11Source ↗PDF ↗Scholar ↗
2021

On Lipschitz Regularization of Convolutional Layers using Toeplitz Matrix Theory

Alexandre Araujo, Benjamin Negrevergne, Yann Chevaleyre, Jamal Atif
AAAI 2021technical

This paper tackles the problem of Lipschitz regularization of Convolutional Neural Networks. Lipschitz regularity is now established as a key property of modern deep learning with implications in training stability, generalization, robustness against adversarial examples, etc. However, computing the…

Cited by 36Source ↗PDF ↗code ↗Scholar ↗