United States

Chain of thought (CoT) is a reasoning framework that can enhance the performance of large language models (LLMs) on complex inference tasks. In particular, among various studies related to CoT, multi-path inference stands out as a simple yet effective improvement. However, there is no optimal setting for the number of inference paths. Therefore, we have to increase the number of inference paths to obtain better results, which in turn increases the inference cost. To address this limitation, we can utilize question-related role templates to guide LLMs into relevant roles, thereby increasing the possibility of correct inferences for each path and further reducing dependence on the number of inference paths while improving reasoning accuracy. However, placing LLMs into specific roles may reduce their reasoning diversity and performance on a few tasks where role dependence is low. To alleviate the excessive immersion of the LLM into a specific role, we propose Nash CoT by constructing a competitive system on each path that balances the generation from role-specific LLMs&#39; and the general LLMs&#39; generation, thereby ensuring both effective role adoption and diversity in LLM generation further maintaining the performance of multi-path inference while reducing the requirement of the number of inference paths. We evaluate Nash CoT across various inference tasks, including Arabic Reasoning, Commonsense Question Answering, and Symbolic Inference, achieving results that are comparable to or better than those of multi-path CoT with the equal number of inference paths.

EMNLP 2024

Nash CoT: Multi-Path Inference with Preference Equilibrium

low-dimensional projected attention

low-rank pre-training

### Welcome to The 63rd Annual Meeting of the Association for Computational Linguistics (ACL 2025)

Message from the General Chair: 
*It is my great pleasure and honor to welcome you to the 63rd Annual Meeting of the Association for Computational Linguistics (ACL 2025), held in beautiful Vienna, Austria, from July 27 to August 1, 2025. ACL2025continues our field’s tradition of excellence in scholarship, innovation, and inclusivity, and I am deeply grateful to the many volunteers who have worked tirelessly to bring this event to life.* 
[Read more](https://drive.google.com/file/d/1GI_hvOpjswAuYdUTromfeDiPpCcqidwg/view?usp=sharing)

To access this event page, you need to log in with the **email address you registered with**. Access credentials will be sent to your email from Underline - subject line "Welcome to ACL 2025". Please be sure to check your spam email folder if you do not see an email confirmation right away.

Please log in

To access this event page, you need to log in with the **email address you registered with**. 

ACL 2025

Welcome to The 63rd Annual Meeting of the Association for Computational Linguistics (ACL 2025)

Chain of thought (CoT) is a reasoning framework that can enhance the performance of large language models (LLMs) on complex inference tasks. In particular, among various studies related to CoT, multi-path inference stands out as a simple yet effective improvement. However, there is no optimal setting for the number of inference paths. Therefore, we have to increase the number of inference paths to obtain better results, which in turn increases the inference cost. To address this limitation, we can utilize question-related role templates to guide LLMs into relevant roles, thereby increasing the possibility of correct inferences for each path and further reducing dependence on the number of inference paths while improving reasoning accuracy. However, placing LLMs into specific roles may reduce their reasoning diversity and performance on a few tasks where role dependence is low. To alleviate the excessive immersion of the LLM into a specific role, we propose Nash CoT by constructing a competitive system on each path that balances the generation from role-specific LLMs' and the general LLMs' generation, thereby ensuring both effective role adoption and diversity in LLM generation further maintaining the performance of multi-path inference while reducing the requirement of the number of inference paths. We evaluate Nash CoT across various inference tasks, including Arabic Reasoning, Commonsense Question Answering, and Symbolic Inference, achieving results that are comparable to or better than those of multi-path CoT with the equal number of inference paths.

poster

## 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.

BERT and TFIDF features excel in capturing rich semantics and important words, respectively. Since most existing clustering methods are solely based on the BERT model, they often fall short in utilizing keyword information, which, however, is very useful in clustering short texts. In this paper, we propose a **CO**-**T**raining **C**lustering (**COTC**) framework to make use of the collective strengths of BERT and TFIDF features. Specifically, we develop two modules responsible for the clustering of BERT and TFIDF features, respectively. We use the deep representations and cluster assignments from the TFIDF module outputs to guide the learning of the BERT module, seeking to align them at both the representation and cluster levels. Reversely, we also use the BERT module outputs to train the TFIDF module, thus leading to the mutual promotion. We then show that the alternating co-training framework can be placed under a unified joint training objective, which allows the two modules to be connected tightly and the training signals to be propagated efficiently. Experiments on eight benchmark datasets show that our method outperforms current SOTA methods significantly.

Leveraging BERT and TFIDF Features for Short Text Clustering via Alignment-Promoting Co-Training

Supervised fine-tuning (SFT) is crucial for adapting Large Language Models (LLMs) to specific tasks. In this work, we demonstrate that the order of training data can lead to significant training imbalances, potentially resulting in performance degradation. Consequently, we propose to mitigate this imbalance by merging SFT models fine-tuned with different data orders, thereby enhancing the overall effectiveness of SFT. Additionally, we introduce a novel technique, "parameter-selection merging," which outperforms traditional weighted-average methods on five datasets. Further, through analysis and ablation studies, we validate the effectiveness of our method and identify the sources of performance improvements.

Mitigating Training Imbalance in LLM Fine-Tuning via Selective Parameter Merging

Large Language Models (LLMs) are capable of producing content that perpetuates stereotypes, discrimination, and toxicity.
The recently proposed \textit{moral self-correction} is a computationally efficient method for reducing harmful content in the responses of LLMs. However, the process of how injecting self-correction instructions can modify the behavior of LLMs remains under-explored. 
In this paper, we explore the effectiveness of moral self-correction by answering three research questions: (1) In what scenarios does moral self-correction work? (2) What are the internal mechanisms of LLMs, e.g., hidden states, that are influenced by moral self-correction instructions? (3) Is intrinsic moral self-correction actually superficial in terms of reduced immorality in hidden states? 

We argue that self-correction can help LLMs find a shortcut to more morally correct output, rather than truly reducing the immorality stored in hidden states.Through empirical investigation with tasks of language generation and multi-choice question answering, we conclude: (i) LLMs exhibit good performance across both tasks, and self-correction instructions are particularly beneficial when the correct answer is already top-ranked; (ii) The morality levels in intermediate hidden states are strong indicators as to whether one instruction would be more effective than another; (iii) Based on our analysis of intermediate hidden states and task case studies of self-correction behaviors, we are first to propose the hypothesis that intrinsic moral self-correction is in fact superficial.

Intrinsic Self-correction for Enhanced Morality: An Analysis of Internal Mechanisms and the Superficial Hypothesis

With large language models (LLMs) achieving remarkable breakthroughs in natural language processing (NLP) domains, recent researchers have actively explored the potential of LLMs for recommendation systems by converting the input data into textual sentences through prompt templates. Although semantic knowledge from LLMs can help enrich the content information of items, to date it is still hard for them to achieve comparable performance to traditional deep learning recommendation models, partly due to a lack of ability to leverage collaborative filtering. In this paper, we propose a novel training-free prompting framework, PepRec, which aims to capture knowledge from both content-based filtering and collaborative filtering to boost recommendation performance with LLMs, while providing interpretation for the recommendation. Experiments based on two real-world datasets from different domains show that PepRec significantly outperforms various traditional deep learning recommendation models and prompt-based recommendation systems.

PepRec: Progressive Enhancement of Prompting for Recommendation

There are recent efforts to $\text{``personalize''}$ large language models (LLMs) by assigning them specific personas. This paper explores the writing styles of such persona-assigned LLMs across different socio-demographic groups based on age, profession, location, and political affiliations, using three widely-used LLMs. Leveraging an existing style embedding model that produces detailed style attributes and latent Dirichlet allocation (LDA) for broad style analysis, we measure style differences using Kullback-Leibler divergence to compare LLM-generated and human-written texts. We find significant style differences among personas. This analysis emphasizes the need to consider socio-demographic factors in language modeling to accurately capture diverse writing styles used for communications. The findings also reveal the strengths and limitations of personalized LLMs, their potential uses, and the importance of addressing biases in their design.

An Empirical Analysis of the Writing Styles of Persona-Assigned LLMs

Communities on social media display distinct patterns of linguistic expression and behaviour, collectively referred to as practices. These practices can be traced in textual exchanges, and reflect the intentions, knowledge, values, and norms of users and communities. This paper introduces a comprehensive methodological workflow for computational identification of such practices within social media texts. By focusing on supporters of Ukraine during the Russia-Ukraine war in (1) the activist collective NAFO and (2) the Eurovision Twitter community, we present a gold-standard data set capturing their unique practices. Using this corpus, we perform practice prediction experiments with both open-source baseline models and OpenAI's large language models (LLMs). Our results demonstrate that closed-source models, especially GPT-4, achieve superior performance, particularly with prompts that incorporate salient features of practices, or utilize Chain-of-Thought prompting. This study provides a detailed error analysis and offers valuable insights into improving the precision of practice identification, thereby supporting context-sensitive moderation and advancing the understanding of online community dynamics.

Detecting Online Community Practices with Large Language Models: A Case Study of Pro-Ukrainian Publics on Twitter

Compositional languages leverage rules that derive meaning from combinations of simpler constituents. This property is considered to be the hallmark of human language as it enables the ability to express novel concepts and ease of learning. As such, numerous studies in the emergent communication field explore the prerequisite conditions for emergence of compositionality. Most of these studies set out one-to-one communication environment wherein a speaker interacts with a single listener during a single round of communication game. However, real-world communications often involve multiple listeners; their interests may vary and they may even need to coordinate among themselves to be successful at a given task. This work investigates the effects of one-to-many communication environment on emergent languages where a single speaker broadcasts its message to multiple listeners to cooperatively solve a task. We observe that simply broadcasting the speaker's message to multiple listeners does not induce more compositional languages. We then find and analyze two axes of environmental pressures that facilitate emergence of compositionality: listeners of *different interests* and *coordination* among listeners.

One-to-Many Communication and Compositionality in Emergent Communication

Sentence transformers excel at grouping topically similar texts, but struggle to differentiate opposing viewpoints on the same topic. This shortcoming hinders their utility in applications where understanding nuanced differences in opinion is essential, such as those related to social and political discourse analysis. This paper addresses this issue by fine-tuning sentence transformers with arguments for and against human-generated controversial claims. We demonstrate how our fine-tuned model enhances the utility of sentence transformers for social computing tasks such as opinion mining and stance detection. We elaborate that applying stance-aware sentence transformers to opinion mining is a more computationally efficient and robust approach in comparison to the classic classification-based approaches.

I love pineapple on pizza != I hate pineapple on pizza: Stance-Aware Sentence Transformers for Opinion Mining

Fact-checking claims is a highly laborious task that involves understanding how each factual assertion within the claim relates to a set of trusted source materials. Existing approaches make sample-level predictions but fail to identify the specific aspects of the claim that are troublesome and the specific evidence relied upon. In this paper, we introduce a method and new benchmark for this challenging task. Our method predicts the fine-grained logical relationship of each aspect of the claim from a set of multimodal documents, which include text, image(s), video(s), and audio(s). We also introduce a new benchmark (M3DC) of claims requiring multimodal multidocument reasoning, which we construct using a novel claim synthesis technique. Experiments show that our approach outperforms other models on this challenging task on two benchmarks while providing finer-grained predictions, explanations, and evidence.

M3D: MultiModal MultiDocument Fine-Grained Inconsistency Detection

Interlinear glossed text (IGT) is a popular format in language documentation projects, where each morpheme is labeled with a descriptive annotation. Automating the creation of interlinear glossed text would be desirable to reduce annotator effort and maintain consistency across annotated corpora. Prior research has explored a number of statistical and neural methods for automatically producing IGT. 

As large language models (LLMs) have showed promising results across multilingual tasks, even for rare, endangered languages, it is natural to wonder whether they can be utilized for the task of generating IGT. We explore whether LLMs can be effective at the task of interlinear glossing with in-context learning, without any traditional training. We propose new approaches for selecting examples to provide in-context, observing that targeted selection can significantly improve performance. We find that LLM-based methods beat standard transformer baselines, despite requiring no training at all. These approaches still underperform state-of-the-art supervised systems for the task, but are highly practical for researchers outside of the NLP community, requiring minimal effort to use.

Downloads

Next from EMNLP 2024

Leveraging BERT and TFIDF Features for Short Text Clustering via Alignment-Promoting Co-Training

Similar lecture

CoT-Valve: Length-Compressible Chain-of-Thought Tuning