United States

Large language models (LLMs) have demonstrated considerable proficiency in general natural language processing (NLP) tasks. 
Instruction tuning, a successful paradigm, enhances the ability of LLMs to follow natural language instructions and exhibit robust generalization across general tasks. 
However, these models often encounter performance limitations across multiple tasks due to constrained model capacity. 
Expanding this capacity during the instruction tuning phase poses significant challenges. 
To address this issue, we introduce parameter-efficient sparsity crafting (PESC), which crafts dense models into sparse models using the mixture-of-experts (MoE) architecture. 
PESC integrates adapters into the MoE layers of sparse models, differentiating experts without altering the individual weights within these layers. 
This method significantly reduces computational costs and GPU memory requirements, facilitating model capacity expansion through a minimal parameter increase when guaranteeing the quality of approximation in function space compared to original sparse upcycling. 
Our empirical evaluation demonstrates the effectiveness of the PESC method. 
Using PESC during instruction tuning, our best sparse model outperforms other sparse and dense models and exhibits superior general capabilities compared to GPT-3.5.
Our code is available at https://github.com/wuhy68/Parameter-Efficient-MoE.

EMNLP 2024

Parameter-Efficient Sparsity Crafting from Dense to Mixture-of-Experts for Instruction Tuning on General Tasks

geometric problem solving

mathematical nlp

large language models

### 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)

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.

Large language models (LLMs) can perform complex reasoning by generating intermediate reasoning steps using chain-of-thought prompting under zero-shot or few-shot settings. However, zero-shot prompting always encounters low performance, and the superior performance of few-shot prompting hinges on the manual-crafting of task-specific demonstrations one by one. In this paper, we present **RoSE** (**R**easoning with **O**rchestrated **S**treaming **E**xperiences), a general framework for solving reasoning tasks that can self-improve as it answers various reasoning questions. To enable RoSE, we describe an architecture that extends an LLM to store all answered reasoning questions and their reasoning steps in a streaming experience pool and orchestrate helpful questions from the pool to assist itself in answering new questions. To set up a question-aware orchestration mechanism, RoSE first calculates the similarity of each question in the pool with the question to be answered. Since the solution to each question in the experience pool is not always correct, RoSE will sort the questions according to their similarity with the question to be answered, and then uniformly divide them into multiple buckets. It finally extracts one question from each bucket to make the extracted questions more diverse. To make the extracted questions help RoSE answer new questions as much as possible, we introduce two other attributes of uncertainty and complexity for each question. RoSE will preferentially select the questions with low uncertainty and high complexity from each bucket. We evaluate the versatility of RoSE in various complex reasoning tasks and LLMs, such as arithmetic and commonsense reasoning, and find that it can achieve excellent performance without any labeled data and pre-set unlabeled data.

Making Large Language Models Better Reasoners with Orchestrated Streaming Experiences

The cross-cultural adaptation of recipes is an important application of identifying and bridging cultural differences in language. The challenge lies in retaining the essence of the original recipe while also aligning with the writing and dietary habits of the target culture. Information Retrieval (IR) offers a way to address the challenge because it retrieves results from the culinary practices of the target culture while maintaining relevance to the original recipe. We introduce a novel task about cross-cultural recipe retrieval and present a unique Chinese-English cross-cultural recipe retrieval benchmark. Our benchmark is manually annotated under limited resource, utilizing various retrieval models to generate a pool of candidate results for manual annotation. The dataset provides retrieval samples that are culturally adapted but textually diverse, presenting greater challenges. We propose CARROT, a plug-and-play cultural-aware recipe information retrieval framework that incorporates cultural-aware query rewriting and re-ranking methods and evaluate it both on our benchmark and intuitive human judgments. The results show that our framework significantly enhances the preservation of the original recipe and its cultural appropriateness for the target culture. We believe these insights will significantly contribute to future research on cultural adaptation.

Bridging Cultures in the Kitchen: A Framework and Benchmark for Cross-Cultural Recipe Retrieval

The advent of transformers has fueled progress in machine translation. More recently large language models (LLMs) have come to the spotlight thanks to their generality and strong performance in a wide range of language tasks, including translation. Here we show that open-source LLMs perform on par with or better than some state-of-the-art baselines in simultaneous machine translation (SiMT) tasks, zero-shot. We also demonstrate that injection of minimal background information, which is easy with an LLM, brings further performance gains, especially on challenging technical subject-matter. This highlights LLMs' potential for building next generation of massively multilingual, context-aware and terminologically accurate SiMT systems that require no resource-intensive training or fine-tuning.

LLMs Are Zero-Shot Context-Aware Simultaneous Translators

In-context learning (ICL) is a few-shot learning paradigm that involves learning mappings through input-output pairs and appropriately applying them to new instances. Despite the remarkable ICL capabilities demonstrated by Large Language Models (LLMs), existing works are highly dependent on large-scale labeled support sets, not always feasible in practical scenarios. To refine this approach, we focus primarily on an innovative selective annotation mechanism, which precedes the standard demonstration retrieval. We introduce the Language Model-based Determinant Point Process (LM-DPP) that simultaneously considers the uncertainty and diversity of unlabeled instances for optimal selection. Consequently, this yields a subset for annotation that strikes a trade-off between the two factors. We apply LM-DPP to various language models, including GPT-J, LlaMA, and GPT-3. Experimental results on 9 NLU and 2 Generation datasets demonstrate that LM-DPP can effectively select canonical examples. Further analysis reveals that LLMs benefit most significantly from subsets that are both low uncertainty and high diversity.

Effective Demonstration Annotation for In-Context Learning via Language Model-Based Determinantal Point Process

Social intelligence is essential for understanding complex human expressions and social interactions. While large multimodal models (LMMs) have demonstrated remarkable performance in social intelligence question answering (SIQA), they are still inclined to generate responses relying on language priors and ignoring the relevant context, due to the dominant prevalence of text-based data in the pre-training stage. To interpret the aforementioned language bias of LMMs, we employ a structure causal model and posit that counterfactual reasoning can mitigate the bias by avoiding spurious correlations between LMMs' internal commonsense knowledge and the given context. However, it is costly and challenging to construct multimodal counterfactual samples. To tackle above challenges, we propose an output $\textbf{D}$istribution $\textbf{C}$alibration network with $\textbf{V}$irtual $\textbf{C}$ounterfactual ($\textbf{DCVC}$) data augmentation framework. DCVC devises a novel output distribution calibration network to mitigate the impact of negative language biases while preserving beneficial priors. Perturbations are introduced to the output distributions of LMMs to simulate the distribution shifts from counterfactual manipulations of the context, which is employed to construct counterfactual augmented data virtually. Experiments on multiple datasets demonstrate the effectiveness and generalizability of our proposed method.

Mitigating Language Bias of LMMs in Social Intelligence Understanding with Virtual Counterfactual Calibration

Recently, tool use with LLMs has become one of the primary research topics as it can help LLM generate truthful and helpful responses. Existing studies on tool use with LLMs primarily focus on enhancing the tool-calling ability of LLMs. In practice, like chat assistants, LLMs are also required to align with human values in the context of tool use. Specifically, LLMs should refuse to answer unsafe tool use relevant instructions and insecure tool responses to ensure their reliability and harmlessness. At the same time, LLMs should demonstrate autonomy in tool use to reduce the costs associated with tool calling. To tackle this issue, we first introduce the principle that LLMs should follow in tool use scenarios: H2A. The goal of H2A is to align LLMs with **helpfulness**, **harmlessness**, and **autonomy**. In addition, we propose ToolAlign, a dataset comprising instruction-tuning data and preference data to align LLMs with the H2A principle for tool use. Based on ToolAlign, we develop LLMs by supervised fine-tuning and preference learning, and experimental results demonstrate that the LLMs exhibit remarkable tool-calling capabilities, while also refusing to engage with harmful content, and displaying a high degree of autonomy in tool utilization. The code and datasets are available at: https://github.com/zhiyuanc2001/ToolAlign.

Towards Tool Use Alignment of Large Language Models

Event Causality Identification (ECI) focuses on extracting causal relations between events in texts. Existing methods for ECI primarily rely on causal features and external knowledge. However, these approaches fall short in two dimensions: (1) causal features between events in a text often lack explicit clues, and (2) external knowledge may introduce bias, while specific problems require tailored analyses. To address these issues, we propose SemDI - a simple and effective ${\bf Sem}$antic ${\bf D}$ependency ${\bf I}$nquiry Network for ECI. SemDI captures semantic dependencies within the context using a unified encoder. Then, it utilizes a ${\it Cloze}$ Analyzer to generate a fill-in token based on comprehensive context understanding. Finally, this fill-in token is used to inquire about the causal relation between two events. Extensive experiments demonstrate the effectiveness of SemDI, surpassing state-of-the-art methods on three widely used benchmarks. Code is available at \url{https://github.com/hrlics/SemDI}.

Advancing Event Causality Identification via Heuristic Semantic Dependency Inquiry Network

Cross-domain Named Entity Recognition (CDNER) is crucial for Knowledge Graph (KG) construction and natural language processing (NLP), enabling learning from source to target domains with limited data. Previous studies often rely on manually collected entity-relevant sentences from the web or attempt to bridge the gap between tokens and entity labels across domains. These approaches are time-consuming and inefficient, as these data are often weakly correlated with the target task and require extensive pre-training.
To address these issues, we propose automatically generating task-oriented knowledge (GTOK) using large language models (LLMs), focusing on the reasoning process of entity extraction. Then, we employ task-oriented pre-training (TOPT) to facilitate domain adaptation. Additionally, current cross-domain NER methods often lack explicit explanations for their effectiveness. Therefore, we introduce the concept of information density to better evaluate the model's effectiveness before performing entity recognition.
We conduct systematic experiments and analyses to demonstrate the effectiveness of our proposed approach and the validity of using information density for model evaluation.

Cross-domain NER with Generated Task-Oriented Knowledge: An Empirical Study from Information Density Perspective

Scene Graph Generation (SGG) provides basic language representation of visual scenes, requiring models to grasp complex and diverse semantics between objects. This complexity and diversity in SGG leads to underrepresentation, where parts of triplet labels are rare or even unseen during training, resulting in imprecise predictions. To tackle this, we propose integrating the pretrained Vision-language Models to enhance representation. However, due to the gap between pretraining and SGG, direct inference of pretrained VLMs on SGG leads to severe bias, which stems from the imbalanced predicates distribution in the pretraining language set. To alleviate the bias, we introduce a novel LM Estimation to approximate the unattainable predicates distribution. Finally, we ensemble the debiased VLMs with SGG models to enhance the representation, where we design a certainty-aware indicator to score each sample and dynamically adjust the ensemble weights. Our training-free method effectively addresses the predicates bias in pretrained VLMs, enhances SGG's representation, and significantly improve the performance.

Predicate Debiasing in Vision-Language Models Integration for Scene Graph Generation Enhancement

While large pre-trained visual-language models have shown promising results on traditional visual question answering benchmarks, it is still challenging for them to answer complex VQA problems which requires diverse world knowledge. Motivated by the research of retrieval-augmented generation in the field of natural language processing, we use Dense Passage Retrieval (DPR) to retrieve related knowledge to help the model answer questions. However, DPR conduct retrieving in natural language space, which may not ensure comprehensive acquisition of image information. Thus, the retrieved knowledge is not truly conducive to helping answer the question, affecting the performance of the overall system. To address this issue, we propose a novel framework that leverages the visual-language model to select the key knowledge retrieved by DPR and answer questions. The framework consists of two modules: Selector and Answerer, where both are initialized by the MLLM and parameter-efficiently finetuned by self-bootstrapping: find key knowledge in the retrieved knowledge documents using the Selector, and then use them to finetune the Answerer to predict answers; obtain the pseudo-labels of key knowledge documents based on the predictions of the Answerer and weak supervision labels, and then finetune the Selector to select key knowledge; repeat. Our framework significantly enhances the performance of the baseline on the challenging open-domain Knowledge-based VQA benchmark, OK-VQA, achieving a state-of-the-art accuracy of 62.83%.

Downloads

Next from EMNLP 2024

Making Large Language Models Better Reasoners with Orchestrated Streaming Experiences

Similar lecture

Upcycling Instruction Tuning from Dense to Mixture-of-Experts via Parameter Merging