Various benchmarks have been proposed to test linguistic understanding in pre-trained vision & language (VL) models. Here we build on the existence task from the VALSE benchmark (Parcalabescu et al., 2022) which we use to test models’ understanding of negation, a particularly interesting issue for multimodal models. However, while such VL benchmarks are useful for measuring model performance, they do not reveal anything about the internal processes through which these models arrive at their outputs in such visio-linguistic tasks. We take inspiration from the growing literature on model interpretability to explain the behaviour of VL models on the understanding of negation. Specifically, we approach these questions through an in-depth analysis of the text encoder in CLIP (Radford et al., 2021), a highly influential VL model. We localise parts of the encoder that process negation and analyse the role of attention heads in this task. Our contributions are threefold. We demonstrate how methods from the language model interpretability literature (e.g., causal tracing) can be translated to multimodal models and tasks; we provide concrete insights into how CLIP processes negation on the VALSE existence task; and we highlight inherent limitations in the VALSE dataset as a benchmark for linguistic understanding.

Taxonomies can serve as a vital foundation for several downstream tasks such as information retrieval and question answering, yet manual construction limits coverage and full potential. Automatic taxonomy induction, particularly using deep Reinforcement Learning (RL), is underexplored in Natural Language Processing (NLP). To address this gap, we present TaxoCritic, a novel approach that leverages deep multi-critic RL agents for taxonomy induction while incorporating credit assignment mechanisms. Our system uniquely assesses different sub-actions within the induction process, providing a granular analysis that aids in the precise attribution of credit and blame. We evaluate the effectiveness of multi-critic algorithms in experiments regarding both accuracy and robustness performance in edge identification. By providing a detailed comparison with state-of-the-art models and highlighting the strengths and limitations of our method, we aim to contribute to the ongoing

There is a growing concern regarding the reproducibility of human evaluation studies in NLP. As part of the ReproHum campaign, we conducted a study to assess the reproducibility of a recent human evaluation study in NLP. Specifically, we attempted to reproduce a human evaluation of a novel approach to enhance Role-Oriented Dialogue Summarization by considering the influence of role interactions. Despite our best efforts to adhere to the reported setup, we were unable to reproduce the statistical results as presented in the original paper. While no contradictory evidence was found, our study raises questions about the validity of the reported statistical significance results, and/or the comprehensiveness with which the original study was reported. In this paper, we provide a comprehensive account of our reproduction study, detailing the methodologies employed, data collection, and analysis procedures. We discuss the implications of our findings for the broader issue of reproducibility in NLP research. Our findings serve as a cautionary reminder of the challenges in conducting reproducible human evaluations and prompt further discussions within the NLP community.

We report our efforts in identifying a set of previous human evaluations in NLP that would be suitable for a coordinated study examining what makes human evaluations in NLP more/less reproducible. We present our results and findings, which include that just 13% of papers had (i) sufficiently low barriers to reproduction, and (ii) enough obtainable information, to be considered for reproduction, and that all but one of the experiments we selected for reproduction was discovered to have flaws that made the meaningfulness of conducting a reproduction questionable. As a result, we had to change our coordinated study design from a reproduce approach to a standardise-then-reproduce-twice approach. Our overall (negative) finding that the great majority of human evaluations in NLP is not repeatable and/or not reproducible and/or too flawed to justify reproduction, paints a dire picture, but presents an opportunity for a rethink about how to design and report human evaluations in NLP.

This paper presents our strategy to address the SemEval-2022 Task 3 PreTENS: Presupposed Taxonomies Evaluating Neural Network Semantics. The goal of the task is to identify if a sentence is deemed acceptable or not, depending on the taxonomic relationship that holds between a noun pair contained in the sentence. For sub-task 1—binary classification—we propose an effective way to enhance the robustness and the generalizability of language models for better classification on this downstream task. We design a two-stage fine-tuning procedure on the ELECTRA language model using data augmentation techniques. Rigorous experiments are carried out using multi-task learning and data-enriched fine-tuning. Experimental results demonstrate that our proposed model, UU-Tax, is indeed able to generalize well for our downstream task. For sub-task 2 —regression—we propose a simple classifier that trains on features obtained from Universal Sentence Encoder (USE). In addition to describing the submitted systems, we discuss other experiments that employ pre-trained language models and data augmentation techniques. For both sub-tasks, we perform error analysis to further understand the behaviour of the proposed models. We achieved a global F1Binary score of 91.25% in sub-task 1 and a rho score of 0.221 in sub-task 2.