United States

The textual content that is fed to Large Language Models (LLMs) can sometimes be noisy con- taining misspelled words. We investigate the effect of such mistakes on the performance of 9 language models, with parameters ranging from 0.2B to 13B, in 3 different NLP tasks, namely Natural Language Inference (NLI), Name Entity Recognition (NER), and Intent Classification (IC). We perform our experiments on 6 different languages (English, German, French, Spanish, Hindi, and Turkish) and build a dictionary of real-world noise for them using Wikipedia edit history. We show that the performance gap of the studied models on the clean and noisy test data averaged across all the datasets and languages ranges from 2.3 to 4.3 absolute percentage points. In addition, mT5 models, in general, show more robustness compared to BLOOM, Falcon, and BERT-like models. In particular, mT5 (13B), was the most robust on average overall, across the 3 tasks, and in 4 out of the 6 languages.

EMNLP 2024

Exploring Robustness of Multilingual LLMs on Real-World Noisy Data

multilingual llms

real-world noise

robustness

### Welcome to the 17th Conference of the European Chapter of the Association for Computational Linguistics (EACL) 
Continuing its mission of expanding and involving the science community of all European countries, EACL had selected the Ukrainian community for the 16th EACL, which took place online due to the COVID pandemic. Unfortunately, the ongoing war made the organisation in Kyiv impossible. Considering the importance of physical interaction among researchers, especially after the restrictions imposed by the COVID pandemic, in addition to an online mode, the conference will be held in Dubrovnik, Croatia, from 2 to 6 of May, 2023. The original aim of strengthening the connection with the Ukrainian community will not change as our program will feature a dedicated session and a workshop to highlight work on Ukrainian language technologies. As the flagship European conference in the field of computational linguistics, EACL welcomes European and international researchers covering a broad spectrum of research areas that are concerned with computational approaches to natural language.

You need to register for the conference in order to access this site. Please visit https://2023.eacl.org/registration for more information.

EACL 2023

As the flagship European conference in the field of computational linguistics, EACL welcomes European and international researchers covering a broad spectrum of research areas that are concerned with computational approaches to natural language.

## Welcome to EMNLP 2024! 
We are excited to welcome you to one of the most prominent conferences in the field of Natural Language Processing. This year, EMNLP 2024 is being held in a hybrid format,
offering both virtual and in-person participation in beautiful Miami. Due to a record-breaking number of submissions, we've expanded the total number of accepted papers to accommodate more cutting-edge research from around the globe.
### [Conference Handbook](https://drive.google.com/file/d/1WPROgxjLAC96AJL7Ugy0tEnYm7dkrbHt/view?usp=sharing)

You are required to register for this event. **Please register [here](https://2024.emnlp.org/registration/).** The EMNLP 2024 event page on Underline will be open to public one week prior to the event.

Please register!

EMNLP 2024 will take place in Miami, Florida from Nov 12th to Nov 16th, 2024, at the Hyatt Regency Miami Hote and on Underline for remote participants.

Large language models (LLMs) have significantly improved their reasoning and decision-making capabilities, as seen in methods like ReAct. However, despite its effectiveness in tackling complex tasks, ReAct faces two main challenges: losing focus on the original question and becoming stuck in action loops. To address these issues, we introduce Focused ReAct, an enhanced version of the ReAct paradigm that incorporates reiteration and early stop mechanisms. These improvements help the model stay focused on the original query and avoid repetitive behaviors. Experimental results show accuracy gains of 18% to 530% and a runtime reduction of up to 34% compared to the original ReAct method.

Focused ReAct: Improving ReAct through Reiterate and Early Stop

We investigate the knowledge of object af- fordances in pre-trained language models (LMs) and pre-trained Vision-Language mod- els (VLMs). A growing body of literature shows that PTLMs fail inconsistently and non- intuitively, demonstrating a lack of reasoning and grounding. To take a first step toward quan- tifying the effect of grounding (or lack thereof), we curate a novel and comprehensive dataset of object affordances – TEXT2AFFORD, charac- terized by 15 affordance classes. Unlike affor- dance datasets collected in vision and language domains, we annotate in-the-wild sentences with objects and affordances. Experimental results reveal that PTLMs exhibit limited rea- soning abilities when it comes to uncommon object affordances. We also observe that pre- trained VLMs do not necessarily capture ob- ject affordances effectively. Through few-shot fine-tuning, we demonstrate improvement in affordance knowledge in PTLMs and VLMs. Our research contributes a novel dataset for lan- guage grounding tasks, and prese

Text2Afford: Probing Object Affordance Prediction abilities of Language Models solely from Text

To predict upcoming text, language models must in some cases retrieve in-context information verbatim. In this report, we investigated how the ability of language models to retrieve arbitrary in-context nouns developed during training (across time) and as language models trained on the same dataset increase in size (across scale). We then asked whether learning of in-context retrieval correlates with learning of more challenging zero-shot benchmarks. Furthermore, inspired by semantic effects in human short-term memory, we evaluated the retrieval with respect to a major semantic component of target nouns, namely whether they denote a concrete or abstract entity, as rated by humans. We show that verbatim in-context retrieval developed in a sudden transition early in the training process, after about 1% of the training tokens. This was observed across model sizes (from 14M and up to 12B parameters), and the transition occurred slightly later for the two smallest models. We further found that the development of verbatim in-context retrieval is positively correlated with the learning of zeroshot benchmarks. Around the transition point, all models showed the advantage of retrieving concrete nouns as opposed to abstract nouns. In all but two smallest models, the advantage dissipated away toward the end of training.

Transformer verbatim in-context retrieval across time and scale

Interpretability studies have played an important role in the field of NLP. They focus on the problems of how models encode information or, for instance, whether linguistic capabilities allow them to prefer grammatical sentences to ungrammatical. Recently, several studies examined whether the models demonstrate patterns similar to humans and whether they are sensitive to the phenomena of interference like humans' grammaticality judgements, including the phenomenon of agreement attraction. In this paper, we probe BERT and GPT models on the syntactic phenomenon of agreement attraction in Russian using the psycholinguistic data with syncretism. Working on the language with syncretism between some plural and singular forms allows us to differentiate between the effects of the surface form and of the underlying grammatical feature. Thus we can further investigate models' sensitivity to this phenomenon and examine if the patterns of their behaviour are similar to human patterns. Moreover, we suggest a new way of comparing models' and humans' responses via statistical testing. We show that there are some similarities between models' and humans' results, while GPT is somewhat more aligned with human responses than BERT. Finally, preliminary results suggest that surface form syncretism influences attraction, perhaps more so than grammatical form syncretism.

Of Models and Men: Probing Neural Networks for Agreement Attraction with Psycholinguistic Data

The ability to compare by analogy, metaphorically or not, lies at the core of how humans understand the world and communicate. In this paper, we study the likelihood of metaphoric outputs, and the capability of a wide range of pretrained transformer-based language models to identify metaphors from other types of analogies, including anomalous ones. In particular, we are interested in discovering whether language models recognise metaphorical analogies equally well as other types of analogies, and whether the model size has an impact on this ability. The results show that there are relevant differences using perplexity as a proxy, with the larger models reducing the gap when it comes to analogical processing, and for distinguishing metaphors from incorrect analogies. This behaviour does not result in increased difficulties for larger generative models in identifying metaphors in comparison to other types of analogies from anomalous sentences in a zero-shot generation setting, when perplexity values of metaphoric and non-metaphoric analogies are similar.

How Are Metaphors Processed by Language Models? The Case of Analogies

Stories are central for interpreting experiences, communicating, and influencing each other via films, medical, media, and other narra tives. Quantifying the similarity between sto ries has numerous applications including detect ing IP infringement, detecting hallucinations, search/recommendation engines, and guiding human-AI collaborations. Despite this, tradi tional NLP text similarity metrics are limited to short text distance metrics like n-gram over laps and embeddings. Larger texts require pre processing with significant information loss through paraphrasing or multi-step decomposi tion. This paper introduces AIStorySimilarity, a novel benchmark to measure the semantic dis tance between long-text stories based on core structural elements drawn from narrative the ory and script writing. Based on four narrative elements (characters, plot, setting, and themes) as well 31 sub-features within these, we use a SOTA LLM (gpt-3.5-turbo) to extract and evaluate the semantic similarity of a diverse set of major Hollywood movies. In addition, we compare human evaluation with story simi larity scores computed three ways: extracting elements from film scripts before evaluation (Elements), directly evaluating entire scripts (Scripts), and extracting narrative elements from the parametric memory of SOTA LLMs without any provided scripts (GenAI). To the best of our knowledge, AIStorySimilarity is the first benchmark to measure long-text story similarity using a comprehensive approach to narrative theory. All code, data, and plot image f iles are available at https://github.com/ jon-chun/AIStorySimiliarity.

AIStorySimilarity: Quantifying Story Similarity Using Narrative for Search, IP Infringement, and Guided Creativity

Sentiment analysis is crucial in Natural Lan guage Processing as it enables the extraction of opinions and emotions from text. How ever, Arabic sentiment analysis is often over looked. Current benchmarks for Arabic senti ment analysis tend to be outdated or lack com prehensive annotations, which limits the devel opment of more accurate and reliable models for the Arabic language. To address these chal lenges, we introduce ArSen, a meticulously an notated Arabic dataset centered on COVID-19, along with IFDHN, a novel model that employs fuzzy logic for more precise sentiment classifi cation1. ArSen offers a robust and contempo rary benchmark, and IFDHN achieves state-of the-art performance in Arabic sentiment analy sis, with 78.12% accuracy, an F1-Macro score of 55.83%, and an F1-Micro score of 78.12% on the test set. Notably, by using only 0.23% of the computational resources of large language models, IFDHN achieved performance compa rable to LLaMA-3-8B, showcasing significant improvements over existing methods.

Advancing Arabic Sentiment Analysis: ArSen Benchmark and the Improved Fuzzy Deep Hybrid Network - Yang Fang, Cheng Xu, Shuhao Guan, Nan Yan, Yuke Mei

The development of Large Language Models (LLMs) has brought impressive performances on mitigation strategies against misinformation, such as counterargument generation. However, LLMs are still seriously hindered by outdated knowledge and by their tendency to generate hallucinated content. In order to circumvent these issues, we propose a new task, namely, Critical Questions Generation, consisting of processing an argumentative text to generate the critical questions (CQs) raised by it. In argumentation theory CQs are tools designed to lay bare the blind spots of an argument by pointing at the information it could be miss- ing. Thus, instead of trying to deploy LLMs to produce knowledgeable and relevant counterar- guments, we use them to question arguments, without requiring any external knowledge. Re- search on CQs Generation using LLMs requires a reference dataset for large scale experimenta- tion. Thus, in this work we investigate two com- plementary methods to create such a resource: (i) instantiating CQs templates as defined by Walton’s argumentation theory and (ii), using LLMs as CQs generators. By doing so, we contribute with a procedure to establish what is a valid CQ and conclude that, while LLMs are reasonable CQ generators, they still have a wide margin for improvement in this task.

Critical Questions Generation: Motivation and Challenges

Humans develop their grammars by making structural generalizations from finite input. We ask how filler-gap dependencies (FGDs), which share a structural generalization despite diverse surface forms, might arise from the input. We explicitly control the input to a neural language model (NLM) to uncover whether the model posits a shared representation for FGDs. We show that while NLMs do have success differentiating grammatical from ungrammatical FGDs, they rely on superficial properties of the input, rather than on a shared generalization. Our work highlights the need for specific linguistic inductive biases to model language acquisition.

Generalizations across filler-gap dependencies in neural language models

The data-driven investigation of the extent to which lexicons of different languages align has mostly fallen into one of two categories: colexification-based and distributional. The two approaches are grounded in distinct methodologies, operate on different assumptions, and are used in diverse ways. This raises two important questions: (a) are there settings in which the predictions of the two approaches can be directly compared? and if so, (b) what is the extent of the similarity and what are its determinants? We offer novel operationalizations for the two approaches in a manner that allows for their direct comparison, and conduct a comprehensive analysis on a diverse set of 16 languages. Our analysis is carried out at different levels of granularity. At the word-level, the two methods present different results across the board. However, intriguingly, at the level of semantic domains (e.g., kinship, quantity), the two methods show considerable convergence in their predictions. Our findings also indicate that the distributional methods likely capture a more fine-grained alignment than their counterpart colexification-based methods, and may thus be more suited for settings where fewer languages are evaluated.

Next from EMNLP 2024

Focused ReAct: Improving ReAct through Reiterate and Early Stop

Similar lecture

Robustification of Multilingual Language Models to Real-world Noise in Crosslingual Zero-shot Settings with Robust Contrastive Pretraining