Publications associées à Uncertainty and Surprisal Jointly Deliver the Punchline: Exploiting Incongruity-Based Features for Humor Recognition

Publications associées (32)

Land Cover Mapping From Multiple Complementary Experts Under Heavy Class Imbalance

Devis Tuia, Valérie Zermatten, Javiera Francisca Castillo Navarro

Deep learning has emerged as a promising avenue for automatic mapping, demonstrating high efficacy in land cover categorization through various semantic segmentation models. Nonetheless, the practical deployment of these models encounters important challen ...

Ieee-Inst Electrical Electronics Engineers Inc2024

Driving and suppressing the human language network using large language models

Martin Schrimpf

Transformer models such as GPT generate human-like language and are predictive of human brain responses to language. Here, using functional-MRI-measured brain responses to 1,000 diverse sentences, we first show that a GPT-based encoding model can predict t ...

Nature Portfolio2024

Infusing structured knowledge priors in neural models for sample-efficient symbolic reasoning

Mattia Atzeni

The ability to reason, plan and solve highly abstract problems is a hallmark of human intelligence. Recent advancements in artificial intelligence, propelled by deep neural networks, have revolutionized disciplines like computer vision and natural language ...

EPFL2024

Augmenting large language models with chemistry tools

Andres Camilo Marulanda Bran, Carlo Baldassari, Philippe Schwaller, Oliver Tobias Schilter

Large language models (LLMs) have shown strong performance in tasks across domains but struggle with chemistry-related problems. These models also lack access to external knowledge sources, limiting their usefulness in scientific applications. We introduce ...

Nature Portfolio2024

Toward Reliable Human Pose Forecasting With Uncertainty

Alexandre Massoud Alahi, Saeed Saadatnejad, Taylor Ferdinand Mordan

Recently, there has been an arms race of pose forecasting methods aimed at solving the spatio-temporal task of predicting a sequence of future 3D poses of a person given a sequence of past observed ones. However, the lack of unified benchmarks and limited ...

2024

ZigZag: Universal Sampling-free Uncertainty Estimation Through Two-Step Inference

Minh Hieu Lê, Nikita Durasov

Whereas the ability of deep networks to produce useful predictions on many kinds of data has been amply demonstrated, estimating the reliability of these predictions remains challenging. Sampling approaches such as MC-Dropout and Deep Ensembles have emerge ...

2024

Design of an Open-Loop Pile-Oscillation Program in the CROCUS Reactor

Andreas Pautz, Vincent Pierre Lamirand, Axel Guy Marie Laureau, Thomas Jean-François Ligonnet

As a follow-up to the CEA-EPFL PETALE experimental program on stainless steel nuclear data, the EPFL initiated an open-loop pile-oscillation experimental program in the CROCUS reactor: BLOOM. A reproduction of the critical experiments of PETALE, the progra ...

2024

Post-correction of Historical Text Transcripts with Large Language Models: An Exploratory Study

Frédéric Kaplan, Maud Ehrmann, Matteo Romanello, Sven Nicolas Yoann Najem-Meyer, Emanuela Boros

The quality of automatic transcription of heritage documents, whether from printed, manuscripts or audio sources, has a decisive impact on the ability to search and process historical texts. Although significant progress has been made in text recognition ( ...

Association for Computational Linguistics2024

Federated learning with uncertainty-based client clustering for fleet-wide fault diagnosis

Olga Fink, Hao Lu

Operators from various industries have been pushing the adoption of wireless sensing nodes for industrial monitoring, and such efforts have produced sizeable condition monitoring datasets that can be used to build diagnosis algorithms capable of warning ma ...

2024

Fly-Swat or Cannon? Cost-Effective Language Model Choice via Meta-Modeling

Robert West, Maxime Jean Julien Peyrard, Marija Sakota

Generative language models (LMs) have become omnipresent across data science. For a wide variety of tasks, inputs can be phrased as natural language prompts for an LM, from whose output the solution can then be extracted. LM performance has consistently be ...

Assoc Computing Machinery2024

Text as a Richer Source of Supervision in Semantic Segmentation Tasks

Devis Tuia, Lloyd Haydn Hughes, Javiera Francisca Castillo Navarro, Valérie Zermatten

This paper introduces TACOSS a text-image alignment approach that allows explainable land cover semantic segmentation by directly integrating semantic concepts encoded from texts. TACOSS combines convolutional neural networks for visual feature extraction ...

The Institute of Electrical and Electronics Engineers, Inc2023

Improving Generalization of Pretrained Language Models

Rabeeh Karimi Mahabadi

In this dissertation, we propose multiple methods to improve transfer learning for pretrained language models (PLMs). Broadly, transfer learning is a powerful technique in natural language processing, where a language model is first pre-trained on a data-r ...

EPFL2023

Gradient-based optimisation of the conditional-value-at-risk using the multi-level Monte Carlo method

Fabio Nobile, Sundar Subramaniam Ganesh

In this work, we tackle the problem of minimising the Conditional-Value-at-Risk (CVaR) of output quantities of complex differential models with random input data, using gradient-based approaches in combination with the Multi-Level Monte Carlo (MLMC) method ...

2023

DISCO: Distilling Counterfactuals with Large Language Models

Antoine Bosselut, Zeming Chen

Models trained with counterfactually augmented data learn representations of the causal structure of tasks, enabling robust generalization. However, high-quality counterfactual data is scarce for most tasks and not easily generated at scale. When crowdsour ...

Assoc Computational Linguistics-Acl2023

Modeling Structured Data in Attention-based Models

Alireza Mohammadshahi

Natural language processing has experienced significant improvements with the development of Transformer-based models, which employ self-attention mechanism and pre-training strategies. However, these models still present several obstacles. A notable issue ...

EPFL2023

Monotonicity Reasoning in the Age of Neural Foundation Models

Zeming Chen

The recent advance of large language models (LLMs) demonstrates that these large-scale foundation models achieve remarkable capabilities across a wide range of language tasks and domains. The success of the statistical learning approach challenges our unde ...

Springer2023

Transformer Models for Vision

Jean-Baptiste Francis Marie Juliette Cordonnier

The recent developments of deep learning cover a wide variety of tasks such as image classification, text translation, playing go, and folding proteins. All these successful methods depend on a gradient-based learning algorithm to train a model on massive ...

EPFL2023

CLIP the Gap: A Single Domain Generalization Approach for Object Detection

Mathieu Salzmann, Martin Pierre Engilberge, Vidit Vidit

Single Domain Generalization (SDG) tackles the problem of training a model on a single source domain so that it generalizes to any unseen target domain. While this has been well studied for image classification, the literature on SDG object detection remai ...

Ieee Computer Soc2023

Revisiting Offline Compression: Going Beyond Factorization-based Methods for Transformer Language Models

Karl Aberer, Rémi Philippe Lebret, Mohammadreza Banaei

Recent transformer language models achieve outstanding results in many natural language processing (NLP) tasks. However, their enormous size often makes them impractical on memory-constrained devices, requiring practitioners to compress them to smaller net ...

Assoc Computational Linguistics-Acl2023

Interpretable Representation Learning and Evaluation for Abstractive Summarization

Andreas Thomas Marfurt

Abstractive summarization has seen big improvements in recent years, mostly due to advances in neural language modeling, language model pretraining, and scaling models and datasets. While large language models generate summaries that are fluent, coherent, ...

EPFL2023