2 citations · 4 across the 18 of their papers we have counts for
22 papers
Enhancing Diffusion Language Models with Autoregressive Post-Training Weights
Yiming Qin, Ke Wang, Amel Abdelraheem +2
Diffusion language models (dLLMs) have emerged as a promising alternative to autoregressive (AR) language models, offering flexible token-update orders and parallel decoding. Recen…
Multiclass Speech Classification Under Noise Disparity
Mahdi Amiri, Sayantan Biswas, Mingchi Hou +2
We investigate noise disparity in multi-class speech classification tasks and develop a strategy to prevent classifiers from exploiting class-dependent noise characteristics. Build…
Augmenting Large Audio Language Models with Low-Level Acoustic Features for Dysarthric Speech Detection
Mahdi Amiri, Hatef Otroshi Shahreza, Pascal Frossard +1
Automatic dysarthric speech detection approaches can support traditional clinical diagnosis, which relies on costly and time-consuming evaluation by a speech and language pathologi…
Recursive Scaling in Masked Diffusion Models
Alba Carballo-Castro, Julianna Piskorz, Paulius Rauba +2
Masked diffusion models (MDMs) generate sequences by iteratively refining a partially masked state and committing tokens in parallel. We introduce recursion in MDMs and propose new…
PriFT: Guiding Supervised Fine-Tuning with Pretrained Priors
Ke Wang, Shuangqi Li, Mathieu Salzmann +1
Supervised fine-tuning (SFT) is an efficient approach for downstream task adaptation and often serves as the initialization stage for reinforcement learning (RL), but it can show w…
Unsupervised Disentanglement Without Compromises : How Functional Orthogonality Enforces Identifiability
Mathieu Cyrille Simon, Pascal Frossard, Christophe De Vleeschouwer
This paper explores unsupervised disentangled representation learning from a functional perspective. We define latent concepts as factors that influence observations through locall…
Fixed-Point Masked Generative Modeling
Andrea Miele, Yiming Qin, Alba Carballo-Castro +2
Masked Generative Models (MGMs) enable parallel decoding and achieve strong performance across modalities, but require full-sequence bidirectional transformers at every step, makin…
Model soups need only one ingredient
Alireza Abdollahpoorrostam, Nikolaos Dimitriadis, Adam Hazimeh +1
Fine-tuning large pre-trained models on a target distribution often improves in-distribution (ID) accuracy, but at the cost of out-of-distribution (OOD) robustness as representatio…
Task Addition and Weight Disentanglement in Closed-Vocabulary Models
Adam Hazimeh, Alessandro Favero, Pascal Frossard
Task arithmetic has recently emerged as a promising method for editing pre-trained \textit{open-vocabulary} models, offering a cost-effective alternative to standard multi-task fin…
Semantic Document Derendering: SVG Reconstruction via Vision-Language Modeling
Adam Hazimeh, Ke Wang, Mark Collier +3
Multimedia documents such as slide presentations and posters are designed to be interactive and easy to modify. Yet, they are often distributed in a static raster format, which lim…
Backdoor Unlearning by Linear Task Decomposition
Amel Abdelraheem, Alessandro Favero, Gerome Bovet +1
Foundation models have revolutionized computer vision by enabling broad generalization across diverse tasks. Yet, they remain highly susceptible to adversarial perturbations and ta…
MEMOIR: Lifelong Model Editing with Minimal Overwrite and Informed Retention for LLMs
Ke Wang, Yiming Qin, Nikolaos Dimitriadis +2
Language models deployed in real-world systems often require post-hoc updates to incorporate new or corrected knowledge. However, editing such models efficiently and reliably-witho…
Single-Input Multi-Output Model Merging: Leveraging Foundation Models for Dense Multi-Task Learning
Juan Garcia Giraldo, Nikolaos Dimitriadis, Ke Wang +1
Model merging is a flexible and computationally tractable approach to merge single-task checkpoints into a multi-task model. Prior work has solely focused on constrained multi-task…
How Compositional Generalization and Creativity Improve as Diffusion Models are Trained
Alessandro Favero, Antonio Sclocchi, Francesco Cagnetta +2
Natural data is often organized as a hierarchical composition of features. How many samples do generative models need in order to learn the composition rules, so as to produce a co…
LiNeS: Post-training Layer Scaling Prevents Forgetting and Enhances Model Merging
Ke Wang, Nikolaos Dimitriadis, Alessandro Favero +3
Fine-tuning pre-trained models has become the standard approach to endow them with specialized knowledge, but it poses fundamental challenges. In particular, \textit{(i)} fine-tuni…
Sequential Representation Learning via Static-Dynamic Conditional Disentanglement
Mathieu Cyrille Simon, Pascal Frossard, Christophe De Vleeschouwer
This paper explores self-supervised disentangled representation learning within sequential data, focusing on separating time-independent and time-varying factors in videos. We prop…
Pareto Low-Rank Adapters: Efficient Multi-Task Learning with Preferences
Nikolaos Dimitriadis, Pascal Frossard, Francois Fleuret
Multi-task trade-offs in machine learning can be addressed via Pareto Front Learning (PFL) methods that parameterize the Pareto Front (PF) with a single model. PFL permits to selec…
Localizing Task Information for Improved Model Merging and Compression
Ke Wang, Nikolaos Dimitriadis, Guillermo Ortiz-Jimenez +2
Model merging and task arithmetic have emerged as promising scalable approaches to merge multiple single-task checkpoints to one multi-task model, but their applicability is reduce…
PUMA: margin-based data pruning
Javier Maroto, Pascal Frossard
Deep learning has been able to outperform humans in terms of classification accuracy in many tasks. However, to achieve robustness to adversarial perturbations, the best methodolog…
IS-Fusion: Instance-Scene Collaborative Fusion for Multimodal 3D Object Detection
Junbo Yin, Jianbing Shen, Runnan Chen +4
Bird's eye view (BEV) representation has emerged as a dominant solution for describing 3D space in autonomous driving scenarios. However, objects in the BEV representation typicall…
Distributional Reduction: Unifying Dimensionality Reduction and Clustering with Gromov-Wasserstein
Hugues Van Assel, Cédric Vincent-Cuaz, Nicolas Courty +3
Unsupervised learning aims to capture the underlying structure of potentially large and high-dimensional datasets. Traditionally, this involves using dimensionality reduction (DR)…
Pi-DUAL: Using Privileged Information to Distinguish Clean from Noisy Labels
Ke Wang, Guillermo Ortiz-Jimenez, Rodolphe Jenatton +3
Label noise is a pervasive problem in deep learning that often compromises the generalization performance of trained models. Recently, leveraging privileged information (PI) -- inf…