Publications

Retreever: Tree-Based Coarse-to-Fine Representations for Retrieval
Shubham Gupta
Zichao Li
Tianyi Chen
Cem Subakan
Siva Reddy
Perouz Taslakian
Valentina Zantedeschi
Document retrieval is a core component of question-answering systems, as it enables conditioning answer generation on new and large-scale co… (voir plus)rpora. While effective, the standard practice of encoding documents into high-dimensional embeddings for similarity search entails large memory and compute footprints, and also makes it hard to inspect the inner workings of the system. In this paper, we propose a tree-based method for organizing and representing reference documents at various granular levels, which offers the flexibility to balance cost and utility, and eases the inspection of the corpus content and retrieval operations. Our method, called ReTreever, jointly learns a routing function per internal node of a binary tree such that query and reference documents are assigned to similar tree branches, hence directly optimizing for retrieval performance. Our evaluations show that ReTreever generally preserves full representation accuracy. Its hierarchical structure further provides strong coarse representations and enhances transparency by indirectly learning meaningful semantic groupings. Among hierarchical retrieval methods, ReTreever achieves the best retrieval accuracy at the lowest latency, proving that this family of techniques can be viable in practical applications.
2024-12-31
arXiv (prépublication)
doi.org
arxiv.org
Revisiting Data Augmentation for Ultrasound Images
Adam Tupper
Christian Gagné
Data augmentation is a widely used and effective technique to improve the generalization performance of deep neural networks. Yet, despite o… (voir plus)ften facing limited data availability when working with medical images, it is frequently underutilized. This appears to come from a gap in our collective understanding of the efficacy of different augmentation techniques across different tasks and modalities. One modality where this is especially true is ultrasound imaging. This work addresses this gap by analyzing the effectiveness of different augmentation techniques at improving model performance across a wide range of ultrasound image analysis tasks. To achieve this, we introduce a new standardized benchmark of 14 ultrasound image classification and semantic segmentation tasks from 10 different sources and covering 11 body regions. Our results demonstrate that many of the augmentations commonly used for tasks on natural images are also effective on ultrasound images, even more so than augmentations developed specifically for ultrasound images in some cases. We also show that diverse augmentation using TrivialAugment, which is widely used for natural images, is also effective for ultrasound images. Moreover, our proposed methodology represents a structured approach for assessing various data augmentations that can be applied to other contexts and modalities.
2024-12-31
Trans. Mach. Learn. Res. (publié)
doi.org
openreview.net
Do Robot Snakes Dream like Electric Sheep? Investigating the Effects of Architectural Inductive Biases on Hallucination
Jerry Huang
Prasanna Parthasarathi
Mehdi Rezagholizadeh
Boxing Chen
A. Chandar
The growth in prominence of large language models (LLMs) in everyday life can be largely attributed to their generative abilities, yet some … (voir plus)of this is also owed to the risks and costs associated with their use. On one front is their tendency to \textit{hallucinate} false or misleading information, limiting their reliability. On another is the increasing focus on the computational limitations associated with traditional self-attention based LLMs, which has brought about new alternatives, in particular recurrent models, meant to overcome them. Yet it remains uncommon to consider these two concerns simultaneously. Do changes in architecture exacerbate/alleviate existing concerns about hallucinations? Do they affect how and where they occur? Through an extensive evaluation, we study how these architecture-based inductive biases affect the propensity to hallucinate. While hallucination remains a general phenomenon not limited to specific architectures, the situations in which they occur and the ease with which specific types of hallucinations can be induced can significantly differ based on the model architecture. These findings highlight the need for better understanding both these problems in conjunction with each other, as well as consider how to design more universal techniques for handling hallucinations.
2024-12-31
ACL (Findings) (publié)
doi.org
arxiv.org
Robust Fine-Tuning from Non-Robust Pretrained Models: Mitigating Suboptimal Transfer With Adversarial Scheduling
Jonas Ngnawe
Maxime Heuillet
Sabyasachi Sahoo
Yann Pequignot
Ola Ahmad
Audrey Durand
Frédéric Precioso
Christian Gagné
Fine-tuning pretrained models is a standard and effective workflow in modern machine learning. However, robust fine-tuning (RFT), which aims… (voir plus) to simultaneously achieve adaptation to a downstream task and robustness to adversarial examples, remains challenging. Despite the abundance of non-robust pretrained models in open-source repositories, their potential for RFT is less understood. We address this knowledge gap by systematically examining RFT from such non-robust models. Our experiments reveal that fine-tuning non-robust models with a robust objective, even under small perturbations, can lead to poor performance, a phenomenon that we dub \emph{suboptimal transfer}. In challenging scenarios (eg, difficult tasks, high perturbation), the resulting performance can be so low that it may be considered a transfer failure. We find that fine-tuning using a robust objective impedes task adaptation at the beginning of training and eventually prevents optimal transfer. However, we propose a novel heuristic, \emph{Epsilon-Scheduling}, a schedule over perturbation strength used during training that promotes optimal transfer. Additionally, we introduce \emph{expected robustness}, a metric that captures performance across a range of perturbations, providing a more comprehensive evaluation of the accuracy-robustness trade-off for diverse models at test time. Extensive experiments on a wide range of configurations (six pretrained models and five datasets) show that \emph{Epsilon-Scheduling} successfully prevents \emph{suboptimal transfer} and consistently improves expected robustness.
2024-12-31
arXiv (prépublication)
doi.org
arxiv.org
Robust Model Predictive Control for the Optimal Operation of the Indoor Environment of a Cluster of Smart Greenhouses
Ehsan Ghorbani
Ahmed Ouammi
Hanane Dagdougui
Smart greenhouses can be defined as cutting-edge technological systems that efficiently control indoor climate conditions to protect crops f… (voir plus)rom harsh outdoor conditions to increase their productivity. In this article, we developed and implemented a robust model predictive control approach that relies on a recursive state estimation method to cope with the impact of measurement and process signal errors. The aim of this approach is to optimally control the internal environment of intelligent greenhouses. A feedback policy problem is decomposing signals for the accessibility of uncertainties. Then, a robust feasibility set can be defined by determining the ellipsoid set on uncertainty to obtain solvable constrained optimization in the CPLEX solver. In the overall formulation, each greenhouse is considered as an independent element. This method can improve the quality of set-point tracking while reducing the computation time required to arrive at a solution. Extensive numerical simulations involving the application of an innovative and robust algorithm to a cluster of greenhouses were conducted to demonstrate the algorithm’s performance and effectiveness.
2024-12-31
IEEE Transactions on AgriFood Electronics (publié)
doi.org
Robust prior-biased acquisition function for human-in-the-loop Bayesian optimization.
Rose Guay-Hottin
Lison Kardassevitch
Hugo Pham
Guillaume Lajoie
Marco Bonizzato
2024-12-31
Knowl. Based Syst. (publié)
doi.org
Rootlets-based registration to the PAM50 spinal cord template
Sandrine Bédard
Jan Valosek
Valeria Oliva
Kenneth A. Weber II
Julien Cohen-Adad
Spinal cord functional MRI studies require precise localization of spinal levels for reliable voxelwise group analyses. Traditional template… (voir plus)-based registration of the spinal cord uses intervertebral discs for alignment. However, substantial anatomical variability across individuals exists between vertebral and spinal levels. This study proposes a novel registration approach that leverages spinal nerve rootlets to improve alignment accuracy and reproducibility across individuals. We developed a registration method leveraging dorsal cervical rootlets segmentation and aligning them non-linearly with the PAM50 spinal cord template. Validation was performed on a multi-subject, multi-site dataset (n=267, 44 sites) and a multi-subject dataset with various neck positions (n=10, 3 sessions). We further validated the method on task-based functional MRI (n=23) to compare group-level activation maps using rootlet-based registration to traditional disc-based methods. Rootlet-based registration showed superior alignment across individuals compared to the traditional disc-based method. Notably, rootlet positions were more stable across neck positions. Group-level analysis of task-based functional MRI using rootlet-based increased Z scores and activation cluster size compared to disc-based registration (number of active voxels from 3292 to 7978). Rootlet-based registration enhances both inter- and intra-subject anatomical alignment and yields better spatial normalization for group-level fMRI analyses. Our findings highlight the potential of rootlet-based registration to improve the precision and reliability of spinal cord neuroimaging group analysis.
2024-12-31
Imaging Neuroscience (publié)
doi.org
arxiv.org
Seeing the world as animals do: How to leverage generative AI for ecological neuroscience
Shahab Bakhtiari
2024-12-31
The Transmitter (publié)
doi.org
Self-Refined Generative Foundation Models for Wireless Traffic Prediction
Chengming Hu
Hao Zhou
Di Wu
Xi Chen
Jun Yan
Xue Liu
With a broad range of emerging applications in 6G networks, wireless traffic prediction has become a critical component of network managemen… (voir plus)t. However, the dynamically shifting distribution of wireless traffic in non-stationary 6G networks presents significant challenges to achieving accurate and stable predictions. Motivated by recent advancements in Generative AI (GenAI)-enabled 6G networks, this paper proposes a novel self-refined Large Language Model (LLM) for wireless traffic prediction, namely TrafficLLM, through in-context learning without parameter fine-tuning or model training. The proposed TrafficLLM harnesses the powerful few-shot learning abilities of LLMs to enhance the scalability of traffic prediction in dynamically changing wireless environments. Specifically, our proposed TrafficLLM embraces an LLM to iteratively refine its predictions through a three-step process: traffic prediction, feedback generation, and prediction refinement. Initially, the proposed TrafficLLM conducts traffic predictions using task-specific demonstration prompts. Recognizing that LLMs may generate incorrect predictions on the first attempt, this paper designs feedback demonstration prompts to provide multifaceted and valuable feedback related to these initial predictions. The validation scheme is further incorporated to systematically enhance the accuracy of mathematical calculations during the feedback generation process. Following this comprehensive feedback, our proposed TrafficLLM introduces refinement demonstration prompts, enabling the same LLM to further refine its predictions and thereby enhance prediction performance. Evaluations on two realistic datasets demonstrate that the proposed TrafficLLM outperforms LLM-based in-context learning methods, achieving performance improvements of 23.17% and 17.09%, respectively.
2024-12-31
IEEE Transactions on Vehicular Technology (publié)
doi.org
arxiv.org
Sensory multi-brain stimulation enhances dyadic cooperative behavior
Ivo Leiva-Cisterna
Paulo Barraza
Eugenio Rodríguez
Guillaume Dumas
Hyperscanning research suggests that interbrain synchronization supports the regulation of social behavior. However, the evidence is predomi… (voir plus)nantly correlational, leaving a gap for epiphenomenal accounts, where synchrony merely represents concurrent stimulus processing rather than a mechanism relevant to interpersonal interactions. Here, we demonstrate that interbrain synchrony causally drives cooperative success, as evidenced by non-invasive stimulation enhancing coupling and subsequently improving performance in a concurrent interdependent cooperation task. We applied dual-sensory entrainment at 16 Hz and 40 Hz to dyads and compared their performance with non-entrained control dyads performing the same cooperation task. We found that dual stimulation improved interbrain synchrony at the targeted frequencies relative to controls, with 16 Hz entrainment producing the most prominent effect. Strikingly, sensory entrainment facilitated sustained behavioral coupling, allowing partners to maintain coordination over extended periods. Notably, these effects are contingent on improved response coordination, indicating the importance of interbrain coupling for facilitating coordination and demonstrating causally that partner neural attunement is necessary to produce effective joint behavior. Thus, our study supports the concept that interbrain synchrony represents a neural mechanism with functional specificity in social interactions.
2024-12-31
Social Cognitive and Affective Neuroscience (publié)
doi.org
Seth H. Holmes, Fruits frais, corps brisés. Les ouvriers agricoles migrants aux États-Unis
Nicolas Roux
2024-12-31
La Nouvelle Revue du Travail (publié)
doi.org
Is sharing always caring? Entropy, boundaries and the plurality of psychotherapeutic process.
Lena Adel
Ana Gómez-Carrillo
Jonas Mago
Michael Lifshitz
Guillaume Dumas
2024-12-31
Physics of Life Reviews (publié)
doi.org