Lambert Mathias
YOU?
Author Swipe
View article: Investigating the effect of masking and background field removal algorithms on the quality of QSM reconstructions using a realistic numerical head phantom
Investigating the effect of masking and background field removal algorithms on the quality of QSM reconstructions using a realistic numerical head phantom Open
Background field removal (BFR) is an important step in the Quantitative Susceptibility Mapping (QSM) pipeline, enabling the reconstruction of local susceptibility distributions by removing contributions from sources outside the region of i…
View article: Reading Recognition in the Wild
Reading Recognition in the Wild Open
To enable egocentric contextual AI in always-on smart glasses, it is crucial to be able to keep a record of the user's interactions with the world, including during reading. In this paper, we introduce a new task of reading recognition to …
View article: Meta-training with Demonstration Retrieval for Efficient Few-shot Learning
Meta-training with Demonstration Retrieval for Efficient Few-shot Learning Open
Large language models show impressive results on few-shot NLP tasks. However, these models are memory and computation-intensive. Meta-training allows one to leverage smaller models for few-shot generalization in a domain-general and task-a…
View article: Logical Satisfiability of Counterfactuals for Faithful Explanations in NLI
Logical Satisfiability of Counterfactuals for Faithful Explanations in NLI Open
Evaluating an explanation's faithfulness is desired for many reasons such as trust, interpretability and diagnosing the sources of model's errors. In this work, which focuses on the NLI task, we introduce the methodology of Faithfulness-th…
View article: TimelineQA: A Benchmark for Question Answering over Timelines
TimelineQA: A Benchmark for Question Answering over Timelines Open
Lifelogs are descriptions of experiences that a person had during their life. Lifelogs are created by fusing data from the multitude of digital services, such as online photos, maps, shopping and content streaming services. Question answer…
View article: Toward a realistic in silico abdominal phantom for QSM
Toward a realistic in silico abdominal phantom for QSM Open
Purpose QSM outside the brain has recently gained interest, particularly in the abdominal region. However, the absence of reliable ground truths makes difficult to assess reconstruction algorithms, whose quality is already compromised by a…
View article: Meta-training with Demonstration Retrieval for Efficient Few-shot Learning
Meta-training with Demonstration Retrieval for Efficient Few-shot Learning Open
Large language models show impressive results on few-shot NLP tasks. However, these models are memory and computation-intensive. Meta-training allows one to leverage smaller models for few-shot generalization in a domain-general and task-a…
View article: TimelineQA: A Benchmark for Question Answering over Timelines
TimelineQA: A Benchmark for Question Answering over Timelines Open
Lifelogs are descriptions of experiences that a person had during their life. Lifelogs are created by fusing data from the multitude of digital services, such as online photos, maps, shopping and content streaming services. Question answer…
View article: ToKen: Task Decomposition and Knowledge Infusion for Few-Shot Hate Speech Detection
ToKen: Task Decomposition and Knowledge Infusion for Few-Shot Hate Speech Detection Open
Hate speech detection is complex; it relies on commonsense reasoning, knowledge of stereotypes, and an understanding of social nuance that differs from one culture to the next. It is also difficult to collect a large-scale hate speech anno…
View article: Logical Satisfiability of Counterfactuals for Faithful Explanations in NLI
Logical Satisfiability of Counterfactuals for Faithful Explanations in NLI Open
Evaluating an explanation's faithfulness is desired for many reasons such as trust, interpretability and diagnosing the sources of model's errors. In this work, which focuses on the NLI task, we introduce the methodology of Faithfulness-th…
View article: Policy Compliance Detection via Expression Tree Inference
Policy Compliance Detection via Expression Tree Inference Open
Policy Compliance Detection (PCD) is a task we encounter when reasoning over texts, e.g. legal frameworks. Previous work to address PCD relies heavily on modeling the task as a special case of Recognizing Textual Entailment. Entailment is …
View article: Hybrid data fidelity term approach for quantitative susceptibility mapping
Hybrid data fidelity term approach for quantitative susceptibility mapping Open
Purpose Susceptibility maps are usually derived from local magnetic field estimations by minimizing a functional composed of a data consistency term and a regularization term. The data‐consistency term measures the difference between the d…
View article: PERFECT: Prompt-free and Efficient Few-shot Learning with Language Models
PERFECT: Prompt-free and Efficient Few-shot Learning with Language Models Open
Current methods for few-shot fine-tuning of pretrained masked language models (PLMs) require carefully engineered prompts and verbalizers for each new task to convert examples into a cloze-format that the PLM can score. In this work, we pr…
View article: Prompt-free and Efficient Few-shot Learning with Language Models
Prompt-free and Efficient Few-shot Learning with Language Models Open
Rabeeh Karimi Mahabadi, Luke Zettlemoyer, James Henderson, Lambert Mathias, Marzieh Saeidi, Veselin Stoyanov, Majid Yazdani. Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2…
View article: ToKen: Task Decomposition and Knowledge Infusion for Few-Shot Hate Speech Detection
ToKen: Task Decomposition and Knowledge Infusion for Few-Shot Hate Speech Detection Open
Badr AlKhamissi, Faisal Ladhak, Srinivasan Iyer, Veselin Stoyanov, Zornitsa Kozareva, Xian Li, Pascale Fung, Lambert Mathias, Asli Celikyilmaz, Mona Diab. Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processi…
View article: UniPELT: A Unified Framework for Parameter-Efficient Language Model Tuning
UniPELT: A Unified Framework for Parameter-Efficient Language Model Tuning Open
Yuning Mao, Lambert Mathias, Rui Hou, Amjad Almahairi, Hao Ma, Jiawei Han, Scott Yih, Madian Khabsa. Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2022.
View article: UNIREX: A Unified Learning Framework for Language Model Rationale Extraction
UNIREX: A Unified Learning Framework for Language Model Rationale Extraction Open
Aaron Chan, Maziar Sanjabi, Lambert Mathias, Liang Tan, Shaoliang Nie, Xiaochang Peng, Xiang Ren, Hamed Firooz. Proceedings of BigScience Episode #5 -- Workshop on Challenges & Perspectives in Creating Large Language Models. 2022.
View article: UNIREX: A Unified Learning Framework for Language Model Rationale Extraction
UNIREX: A Unified Learning Framework for Language Model Rationale Extraction Open
An extractive rationale explains a language model's (LM's) prediction on a given task instance by highlighting the text inputs that most influenced the prediction. Ideally, rationale extraction should be faithful (reflective of LM's actual…
View article: UniPELT: A Unified Framework for Parameter-Efficient Language Model Tuning
UniPELT: A Unified Framework for Parameter-Efficient Language Model Tuning Open
Recent parameter-efficient language model tuning (PELT) methods manage to match the performance of fine-tuning with much fewer trainable parameters and perform especially well when training data is limited. However, different PELT methods …
View article: Streaking artifact suppression of quantitative susceptibility mapping reconstructions via L1‐norm data fidelity optimization (L1‐QSM)
Streaking artifact suppression of quantitative susceptibility mapping reconstructions via L1‐norm data fidelity optimization (L1‐QSM) Open
Purpose The presence of dipole‐inconsistent data due to substantial noise or artifacts causes streaking artifacts in quantitative susceptibility mapping (QSM) reconstructions. Often used Bayesian approaches rely on regularizers, which in t…
View article: Findings of the WOAH 5 Shared Task on Fine Grained Hateful Memes Detection
Findings of the WOAH 5 Shared Task on Fine Grained Hateful Memes Detection Open
Lambert Mathias, Shaoliang Nie, Aida Mostafazadeh Davani, Douwe Kiela, Vinodkumar Prabhakaran, Bertie Vidgen, Zeerak Waseem. Proceedings of the 5th Workshop on Online Abuse and Harms (WOAH 2021). 2021.
View article: Personalized Query Rewriting in Conversational AI Agents
Personalized Query Rewriting in Conversational AI Agents Open
Spoken language understanding (SLU) systems in conversational AI agents often experience errors in the form of misrecognitions by automatic speech recognition (ASR) or semantic gaps in natural language understanding (NLU). These errors eas…
View article: Machine Learning to predict tuberculosis in cattle from the state of Sao Paulo, Brazil
Machine Learning to predict tuberculosis in cattle from the state of Sao Paulo, Brazil Open
Tuberculosis is a well-known and worldwide spread zoonosis. In Brazil 1.594.787 cases were confirmed cases since 2001, where, in Sao Paulo state, 8.226 deaths were reported. This study aims to present steps related to the use of machine le…
View article: Pre-Training for Query Rewriting in A Spoken Language Understanding System
Pre-Training for Query Rewriting in A Spoken Language Understanding System Open
Query rewriting (QR) is an increasingly important technique to reduce customer friction caused by errors in a spoken language understanding pipeline, where the errors originate from various sources such as speech recognition errors, langua…
View article: Leveraging External Knowledge for Out-Of-Vocabulary Entity Labeling
Leveraging External Knowledge for Out-Of-Vocabulary Entity Labeling Open
Dealing with previously unseen slots is a challenging problem in a real-world multi-domain dialogue state tracking task. Other approaches rely on predefined mappings to generate candidate slot keys, as well as their associated values. This…
View article: Improving Long Distance Slot Carryover in Spoken Dialogue Systems
Improving Long Distance Slot Carryover in Spoken Dialogue Systems Open
Tracking the state of the conversation is a central component in task-oriented spoken dialogue systems. One such approach for tracking the dialogue state is slot carryover, where a model makes a binary decision if a slot from the context i…
View article: A dataset for resolving referring expressions in spoken dialogue via contextual query rewrites (CQR)
A dataset for resolving referring expressions in spoken dialogue via contextual query rewrites (CQR) Open
We present Contextual Query Rewrite (CQR) a dataset for multi-domain task-oriented spoken dialogue systems that is an extension of the Stanford dialog corpus (Eric et al., 2017a). While previous approaches have addressed the issue of diver…
View article: Scaling Multi-Domain Dialogue State Tracking via Query Reformulation
Scaling Multi-Domain Dialogue State Tracking via Query Reformulation Open
We present a novel approach to dialogue state tracking and referring expression resolution tasks. Successful contextual understanding of multi-turn spoken dialogues requires resolving referring expressions across turns and tracking the ent…
View article: Time Masking: Leveraging Temporal Information in Spoken Dialogue Systems
Time Masking: Leveraging Temporal Information in Spoken Dialogue Systems Open
In a spoken dialogue system, dialogue state tracker (DST) components track the state of the conversation by updating a distribution of values associated with each of the slots being tracked for the current user turn, using the interactions…
View article: Scaling Multi-Domain Dialogue State Tracking via Query Reformulation
Scaling Multi-Domain Dialogue State Tracking via Query Reformulation Open
Pushpendre Rastogi, Arpit Gupta, Tongfei Chen, Mathias Lambert. Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 2 (Industry Papers). 201…