55 citations · 209 across the 32 of their papers we have counts for
17 papers · 1 filter
Improved Confidence Estimates for Black-Box Large Language Models
Sokhna Diarra Mbacke, Mouloud Belbahri, Gabriel Loaiza-Ganem
Uncertainty quantification (UQ) is essential for the safe deployment of large language models (LLMs). Existing methods, from verbalized confidence to ones requiring multiple genera…
Conf-Gen: Conformal Uncertainty Quantification for Generative Models
Gabriel Loaiza-Ganem, Kevin Zhang, Wei Cui +2
Conformal prediction (CP) and its extension, conformal risk control (CRC), are established frameworks for quantifying uncertainty in supervised machine learning through formal guar…
On Convolutions, Intrinsic Dimension, and Diffusion Models
Kin Kwan Leung, Rasa Hosseinzadeh, Gabriel Loaiza-Ganem
The manifold hypothesis asserts that data of interest in high-dimensional ambient spaces, such as image data, lies on unknown low-dimensional submanifolds. Diffusion models (DMs) -…
Textual Bayes: Quantifying Prompt Uncertainty in LLM-Based Systems
Brendan Leigh Ross, Noël Vouitsis, Atiyeh Ashari Ghomi +8
Although large language models (LLMs) are becoming increasingly capable of solving challenging real-world tasks, accurately quantifying their uncertainty remains a critical open pr…
Last Layer Empirical Bayes
Valentin Villecroze, Yixin Wang, Gabriel Loaiza-Ganem
The task of quantifying the inherent uncertainty associated with neural network predictions is a key challenge in artificial intelligence. Bayesian neural networks (BNNs) and deep…
Deep Ensembles Secretly Perform Empirical Bayes
Gabriel Loaiza-Ganem, Valentin Villecroze, Yixin Wang
Quantifying uncertainty in neural networks is a highly relevant problem which is essential to many applications. The two predominant paradigms to tackle this task are Bayesian neur…
Inconsistencies In Consistency Models: Better ODE Solving Does Not Imply Better Samples
Noël Vouitsis, Rasa Hosseinzadeh, Brendan Leigh Ross +4
Although diffusion models can generate remarkably high-quality samples, they are intrinsically bottlenecked by their expensive iterative sampling procedure. Consistency models (CMs…
A Geometric View of Data Complexity: Efficient Local Intrinsic Dimension Estimation with Diffusion Models
Hamidreza Kamkari, Brendan Leigh Ross, Rasa Hosseinzadeh +2
High-dimensional data commonly lies on low-dimensional submanifolds, and estimating the local intrinsic dimension (LID) of a datum -- i.e. the dimension of the submanifold it belon…
Deep Generative Models through the Lens of the Manifold Hypothesis: A Survey and New Connections
Gabriel Loaiza-Ganem, Brendan Leigh Ross, Rasa Hosseinzadeh +2
In recent years there has been increased interest in understanding the interplay between deep generative models (DGMs) and the manifold hypothesis. Research in this area focuses on…
A Geometric Explanation of the Likelihood OOD Detection Paradox
Hamidreza Kamkari, Brendan Leigh Ross, Jesse C. Cresswell +3
Likelihood-based deep generative models (DGMs) commonly exhibit a puzzling behaviour: when trained on a relatively complex dataset, they assign higher likelihood values to out-of-d…
Data-Efficient Multimodal Fusion on a Single GPU
Noël Vouitsis, Zhaoyan Liu, Satya Krishna Gorti +5
The goal of multimodal alignment is to learn a single latent space that is shared between multimodal inputs. The most powerful models in this space have been trained using massive…
Exposing flaws of generative model evaluation metrics and their unfair treatment of diffusion models
George Stein, Jesse C. Cresswell, Rasa Hosseinzadeh +7
We systematically study a wide variety of generative models spanning semantically-diverse image datasets to understand and improve the feature extractors and metrics used to evalua…
TR0N: Translator Networks for 0-Shot Plug-and-Play Conditional Generation
Zhaoyan Liu, Noel Vouitsis, Satya Krishna Gorti +2
We propose TR0N, a highly general framework to turn pre-trained unconditional generative models, such as GANs and VAEs, into conditional models. The conditioning can be highly arbi…
Denoising Deep Generative Models
Gabriel Loaiza-Ganem, Brendan Leigh Ross, Luhuan Wu +3
Likelihood-based deep generative models have recently been shown to exhibit pathological behaviour under the manifold hypothesis as a consequence of using high-dimensional densitie…
Relating Regularization and Generalization through the Intrinsic Dimension of Activations
Bradley C. A. Brown, Jordan Juravsky, Anthony L. Caterini +1
Given a pair of models with similar training set performance, it is natural to assume that the model that possesses simpler internal representations would exhibit better generaliza…
Bayesian Nonparametrics for Offline Skill Discovery
Valentin Villecroze, Harry J. Braviner, Panteha Naderian +2
Skills or low-level policies in reinforcement learning are temporally extended actions that can speed up learning and enable complex behaviours. Recent work in offline reinforcemen…
C-Learning: Horizon-Aware Cumulative Accessibility Estimation
Panteha Naderian, Gabriel Loaiza-Ganem, Harry J. Braviner +4
Multi-goal reaching is an important problem in reinforcement learning needed to achieve algorithmic generalization. Despite recent advances in this field, current algorithms suffer…