NLP in the era of generative AI, cognitive sciences, and societal transformation
Join us at Mila in October for a three-day workshop to explore the transformative potential of language technologies and their implications for society.
This program is designed to provide decision-makers, policymakers and professional working in policy with a foundational understanding of AI technology.
We use cookies to analyze the browsing and usage of our website and to personalize your experience. You can disable these technologies at any time, but this may limit certain functionalities of the site. Read our Privacy Policy for more information.
Setting cookies
You can enable and disable the types of cookies you wish to accept. However certain choices you make could affect the services offered on our sites (e.g. suggestions, personalised ads, etc.).
Essential cookies
These cookies are necessary for the operation of the site and cannot be deactivated. (Still active)
Analytics cookies
Do you accept the use of cookies to measure the audience of our sites?
Multimedia Player
Do you accept the use of cookies to display and allow you to watch the video content hosted by our partners (YouTube, etc.)?
Publications
Systematic Generalization by Finetuning? Analyzing Pretrained Language Models Using Constituency Tests
Constituents are groups of words that behave as a syntactic unit. Many linguistic phenomena (e.g., question formation, diathesis alternation… (see more)s) require the manipulation and rearrangement of constituents in a sentence. In this paper, we investigate how different finetuning setups affect the ability of pretrained sequence-to-sequence language models such as BART and T5 to replicate constituency tests — transformations that involve manipulating constituents in a sentence. We design multiple evaluation settings by varying the combinations of constituency tests and sentence types that a model is exposed to during finetuning. We show that models can replicate a linguistic transformation on a specific type of sentence that they saw during finetuning, but performance degrades substantially in other settings, showing a lack of systematic generalization. These results suggest that models often learn to manipulate sentences at a surface level unrelated to the constituent-level syntactic structure, for example by copying the first word of a sentence. These results may partially explain the brittleness of pretrained language models in downstream tasks.
2023-12-01
Proceedings of the 6th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP (published)
Social media platforms present a perplexing duality, acting at once as sites to build community and a sense of belonging, while also giving … (see more)rise to misinformation, facilitating and intensifying disinformation campaigns and perpetuating existing patterns of discrimination from the physical world. The first-step platforms take in mitigating the harmful side of social media involves identifying and managing toxic content. Users produce an enormous volume of posts which must be evaluated very quickly. This is an application context that requires machine-learning (ML) tools, but as we detail in this article, ML approaches rely on human annotators, analysts, and moderators. Our review of existing methods and potential improvements indicates that neither humans nor ML can be removed from this process in the near future. However, we see room for improvement in the working conditions of these human workers.
The landscape of available textual adversarial attacks keeps growing, posing severe threats and raising concerns regarding the deep NLP syst… (see more)em's integrity. However, the crucial problem of defending against malicious attacks has only drawn the attention of the NLP community. The latter is nonetheless instrumental in developing robust and trustworthy systems. This paper makes two important contributions in this line of search: (i) we introduce LAROUSSE, a new framework to detect textual adversarial attacks and (ii) we introduce STAKEOUT, a new benchmark composed of nine popular attack methods, three datasets, and two pre-trained models. LAROUSSE is ready-to-use in production as it is unsupervised, hyperparameter-free, and non-differentiable, protecting it against gradient-based methods. Our new benchmark STAKEOUT allows for a robust evaluation framework: we conduct extensive numerical experiments which demonstrate that LAROUSSE outperforms previous methods, and which allows to identify interesting factors of detection rate variations.
2023-12-01
Findings of the Association for Computational Linguistics: EMNLP 2023 (published)
Validation of an AI-assisted Treatment Outcome Measure for Gender-Affirming Voice Care: Comparing AI Accuracy to Listener's Perception of Voice Femininity.
Validation of an AI-assisted Treatment Outcome Measure for Gender-Affirming Voice Care: Comparing AI Accuracy to Listener's Perception of Voice Femininity.
Data scarcity is a crucial issue for the development of highly multilingual NLP systems. Yet for many under-represented languages (ULs) -- l… (see more)anguages for which NLP re-search is particularly far behind in meeting user needs -- it is feasible to annotate small amounts of data. Motivated by this, we propose XTREME-UP, a benchmark defined by: its focus on the scarce-data scenario rather than zero-shot; its focus on user-centric tasks -- tasks with broad adoption by speakers of high-resource languages; and its focus on under-represented languages where this scarce-data scenario tends to be most realistic. XTREME-UP evaluates the capabilities of language models across 88 under-represented languages over 9 key user-centric technologies including ASR, OCR, MT, and information access tasks that are of general utility. We create new datasets for OCR, autocomplete, semantic parsing, and transliteration, and build on and refine existing datasets for other tasks. XTREME-UP provides methodology for evaluating many modeling scenarios including text-only, multi-modal (vision, audio, and text),supervised parameter tuning, and in-context learning. We evaluate commonly used models on the benchmark. We release all code and scripts to train and evaluate models
2023-12-01
Findings of the Association for Computational Linguistics: EMNLP 2023 (published)
This paper advances the field of pedestrian localization by introducing a unifying framework for opportunistic positioning based on nonlinea… (see more)r factor graph optimization. While many existing approaches assume constant availability of one or multiple sensing signals, our methodology employs IMU-based pedestrian inertial navigation as the backbone for sensor fusion, opportunistically integrating Ultra-Wideband (UWB), Bluetooth Low Energy (BLE), and WiFi signals when they are available in the environment. The proposed PEOPLEx framework is designed to incorporate sensing data as it becomes available, operating without any prior knowledge about the environment (e.g. anchor locations, radio frequency maps, etc.). Our contributions are twofold: 1) we introduce an opportunistic multi-sensor and real-time pedestrian positioning framework fusing the available sensor measurements; 2) we develop novel factors for adaptive scaling and coarse loop closures, significantly improving the precision of indoor positioning. Experimental validation confirms that our approach achieves accurate localization estimates in real indoor scenarios using commercial smartphones.
Chaotic systems such as the gravitational N-body problem are ubiquitous in astronomy. Machine learning (ML) is increasingly deployed to pred… (see more)ict the evolution of such systems, e.g. with the goal of speeding up simulations. Strategies such as active Learning (AL) are a natural choice to optimize ML training. Here we showcase an AL failure when predicting the stability of the Sitnikov three-body problem, the simplest case of N-body problem displaying chaotic behavior. We link this failure to the fractal nature of our classification problem's decision boundary. This is a potential pitfall in optimizing large sets of N-body simulations via AL in the context of star cluster physics, galactic dynamics, or cosmology.
Examining the detailed structure of galaxy populations provides valuable insights into their formation and evolution mechanisms. Significant… (see more) barriers to such analysis are the non-trivial noise properties of real astronomical images and the point spread function (PSF) which blurs structure. Here we present a framework which combines recent advances in score-based likelihood characterization and diffusion model priors to perform a Bayesian analysis of image deconvolution. The method, when applied to minimally processed \emph{Hubble Space Telescope} (\emph{HST}) data, recovers structures which have otherwise only become visible in next-generation \emph{James Webb Space Telescope} (\emph{JWST}) imaging.