Antoine Bosselut

Related publications (26)

About
Privacy
Disclaimer

Graph Chatbot

Let Me Teach You: Pedagogical Foundations of Feedback for Language Models

Antoine Bosselut, Beatriz Maria Borges Ribeiro, Tanja Christina Käser Jacober

Natural Language Feedback (NLF) is an increasingly popular mechanism for aligning Large Language Models (LLMs) to human preferences. Despite the diversity of the information it can convey, NLF methods are often hand-designed and arbitrary, with little syst ...

Association for Computational Linguistics2024

“Flex Tape Can't Fix That”: Bias and Misinformation in Edited Language Models

Antoine Bosselut, Syrielle Mathilde Marie Montariol

Weight-based model editing methods update the parametric knowledge of language models post-training. However, these methods can unintentionally alter unrelated parametric knowledge representations, potentially increasing the risk of harm. In this work, we ...

Association for Computational Linguistics (ACL)2024

Discovering Knowledge-Critical Subnetworks in Pretrained Language Models

Antoine Bosselut, Zeming Chen, Negar Foroutan Eghlidi, Deniz Bayazit

Pretrained language models (LMs) encode implicit representations of knowledge in their parameters. However, localizing these representations and disentangling them from each other remains an open problem. In this work, we investigate whether pretrained lan ...

Association for Computational Linguistics2024

CRAB: Assessing the Strength of Causal Relationships Between Real-World Events

Karl Aberer, Antoine Bosselut, Angeliki Romanou, Debjit Paul, Léo Jules Laugier, Syrielle Mathilde Marie Montariol

Understanding narratives requires reasoning about the cause-and-effect relationships between events mentioned in the text. While existing foundation models yield impressive results in many NLP tasks requiring reasoning, it is unclear whether they understan ...

Association for Computational Linguistics2023

PEACOK: Persona Commonsense Knowledge for Consistent and Engaging Narratives

Antoine Bosselut, Beatriz Maria Borges Ribeiro, Silin Gao, Deniz Bayazit, Soyoung Oh

Sustaining coherent and engaging narratives requires dialogue or storytelling agents to understand how the personas of speakers or listeners ground the narrative. Specifically, these agents must infer personas of their listeners to produce statements that ...

Assoc Computational Linguistics-Acl2023

Breaking the Language Barrier: Improving Cross-Lingual Reasoning with Structured Self-Attention

Karl Aberer, Antoine Bosselut, Negar Foroutan Eghlidi, Mohammadreza Banaei

In this work, we study whether multilingual language models (MultiLMs) can transfer logical reasoning abilities to other languages when they are fine-tuned for reasoning in a different language. We evaluate the cross-lingual reasoning abilities of MultiLMs ...

Association for Computational Linguistics2023

kogito: A Commonsense Knowledge Inference Toolkit

Antoine Bosselut, Mahammad Ismayilzada

In this paper, we present kogito, an open-source tool for generating commonsense inferences about situations described in text. kogito provides an intuitive and extensible interface to interact with natural language generation models that can be used for h ...

Assoc Computational Linguistics-Acl2023

Mitigating Label Biases for In-context Learning

Antoine Bosselut, Zeming Chen

Various design settings for in-context learning (ICL), such as the choice and order of the in-context examples, can bias the model's predictions. While many studies discuss these design choices, there have been few systematic investigations into categorizi ...

Assoc Computational Linguistics-Acl2023

DISCO: Distilling Counterfactuals with Large Language Models

Antoine Bosselut, Zeming Chen, Qiyue Gao

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

REFINER: Reasoning Feedback on Intermediate Representations

Boi Faltings, Robert West, Maxime Jean Julien Peyrard, Antoine Bosselut, Beatriz Maria Borges Ribeiro, Debjit Paul, Mahammad Ismayilzada

Language models (LMs) have recently shown remarkable performance on reasoning tasks by explicitly generating intermediate inferences, e.g., chain-of-thought prompting. However, these intermediate inference steps may be in- appropriate deductions from the i ...

2023

Page 2 of 3