United States

Large language models (LLMs) have transformed natural language processing but pose significant challenges for real-world deployment. These models necessitate considerable computing resources, which can be costly and frequently unavailable. Model compression techniques such as quantization are often leveraged to alleviate resource demand, but they may have a negative impact on the generation quality. In this study, we explore the integration of 4-bit Post-training Quantization (PTQ) with QLoRA to address these issues. We demonstrate through extensive experiments that this integration outperforms standard PTQ, and in some cases even 16-bit full-parameter fine-tuning on LLMs, validated across proprietary and public datasets with different quantization algorithms. The results demonstrate the efficacy of PTQ-QLoRA integration, offering a viable solution for deploying powerful LLMs in resource-constrained environments without compromising on performance.

NAACL 2025

Can Post-Training Quantization Benefit from an Additional QLoRA Integration?

accuracy recovery

qlora

post-training quantization

### 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 2025 Annual Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics

Welcome to the 2025 meeting of the Nations of the Americas Chapter of the Association for Computational Linguistics! I am proud to help organize the first NAACL conference to carry the new name of our organization, one that emphasizes inclusion for all of the Americas. I am also pleased to welcome you to Albuquerque, New Mexico, a state whose unique blend of cultural influences will make for an excellent backdrop for NAACL 2025, especially with this year’s special theme on NLP in a Multicultural World. 
**[Continue reading...](https://drive.google.com/file/d/1jX-qGhqVSZZCIrAnJaz798pflu5Irrdn/view?usp=sharing)**

*- Colin Cherry, Google, NAACL 2025 General Chair* 

[![](https://assets.underline.io/markdown_image/1/image/b087f8a4dc5816d6e1a6514e59c59ac3.png)](https://drive.google.com/file/d/1T96GzPqObXrMln2BMByCSXTSizjTg69P/view?usp=sharing)

You need to log in with the email address you registered with. Access credentials have been sent to your email. 

Please be sure to check your spam and other email folders if you do not see an email confirmation right away.

Please log in to explore this event.

To access NAACL 2025 event page you are required to register. Please follow [**this link**](https://2025.naacl.org/registration/registration/) to register. Access will depend on your registration type.

Please register!

Welcome to 2025 Annual Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics

Large Language Models (LLMs) face limitations in AI legal and policy applications due to outdated knowledge, hallucinations, and poor reasoning in complex contexts. Retrieval-Augmented Generation (RAG) systems address these issues by incorporating external knowledge, but suffer from retrieval errors, ineffective context integration, and high operational costs. This paper presents the Hybrid Parameter-Adaptive RAG (HyPA-RAG) system, designed for the AI legal domain, with NYC Local Law 144 (LL144) as the test case. HyPA-RAG integrates a query complexity classifier for adaptive parameter tuning, a hybrid retrieval approach combining dense, sparse, and knowledge graph methods, and a comprehensive evaluation framework with tailored question types and metrics. Testing on LL144 demonstrates that HyPA-RAG enhances retrieval accuracy, response fidelity, and contextual precision, offering a robust and adaptable solution for high-stakes legal and policy applications.

HyPA-RAG: A Hybrid Parameter Adaptive Retrieval-Augmented Generation System for AI Legal and Policy Applications

In the digital age, social media platforms like Twitter serve as an extensive repository of public discourse, including instances of sexism. It is important to identify such behavior since radicalized ideologies can lead to real-world violent acts. This project aims to create a deep learning-based tool using GraphSAGE, sentiment analysis and natural language processing (NLP) techniques to capture complex contextual information and relationships within text. The tool will be designed to analyze tweets for sexism detection and classify them into five categories.

Detecting Sexism in Tweets: A Sentiment Analysis and Graph Neural Network Approach

Arabic calligraphy carries rich historical information and meaning. However, the complexity of its artistic elements and the absence of a consistent baseline make text extraction from such works highly challenging. In this paper, we provide an in-depth analysis of the unique obstacles in processing and interpreting these images, including the variability in calligraphic styles, the influence of artistic distortions, and the challenges posed by missing or damaged text elements. We explore potential solutions by leveraging state-of-the-art architectures and deep learning models, including visual language models, to improve text extraction and script completion.

Text Extraction and Script Completion in Images of Arabic Script-Based Calligraphy: A Thesis Proposal

Identifying commercial posts in resource-constrained languages among diverse and unstructured content remains a significant challenge for automatic text classification tasks. To address this, this work introduces a novel dataset named MDC3 (Multimodal Dataset for Commercial Content Classification), comprising 5,007 annotated Bengali social media posts classified as commercial and noncommercial. A comprehensive annotation guideline accompanying the dataset is included to aid future dataset creation in resource-constrained languages. Furthermore, we performed extensive experiments on MDC3 considering both unimodal and multimodal domains. Specifically, the late fusion of textual (mBERT) and visual (ViT) models (i.e., ViT+mBERT) achieves the highest F1 score of 90.91, significantly surpassing other baselines.

MDC3: A Novel Multimodal Dataset for Commercial Content Classification in Bengali

LLMs are crucial for generative AI, but their causal reasoning abilities remain uncertain, especially at Pearl’s interventional and counterfactual levels. This study presents a dataset of over 7,000 causal questions in Portuguese, sourced from real human interactions. A subset was annotated using a novel taxonomy based on causal intent, action requirements, and the required level of reasoning. Evaluations with GPT-4o and Llama3.1 reveal significant challenges in advanced causal tasks, highlighting LLMs’ limitations and the need for improvements in causal reasoning for AI agents.

Advancing Causal Reasoning in Portuguese: An Evaluation of LLMs in Interventions and Counterfactuals

Large Language Models (LLMs) have excelled in multi-hop question-answering (M-QA) due to their advanced reasoning abilities. However, the impact of inherent reasoning structures on LLM M-QA performance remains unclear. To fill this gap, we introduce the Graph Reasoning-Structured Question Answering Dataset (GRS-QA), which includes both semantic contexts and reasoning structures for QA pairs. Unlike existing M-QA~datasets, where different reasoning structures are entangled, GRS-QA explicitly captures reasoning pathways through reasoning graphs, where nodes represent textual contexts and edges denote logical flows. These graphs enable fine-grained evaluation of LLM reasoning across various structures. Our empirical analysis shows that LLMs perform differently based on reasoning structures, highlighting the role of textual structure in reasoning beyond semantics.

GRS-QA - Graph Reasoning-Structured Question Answering Dataset

Generating cell-level descriptions for Jupyter Notebooks, which is a major resource consisting of codes, tables, and descriptions, has been attracting increasing research attention. However, existing methods for Jupyter Notebooks mostly focus on generating descriptions from code snippets or table outputs independently. On the other side, descriptions should be personalized as users have different purposes in different scenarios while previous work ignored this situation during description generation. In this work, we formulate a new task, personalized description generation with code, tables, and user-written guidelines in Jupyter Notebooks. To evaluate this new task, we collect and propose a benchmark, namely TOOL , containing code, tables, and user-written guidelines as inputs and personalized descriptions as targets. Extensive experiments show that while existing models of text generation are able to generate fluent and readable descriptions, they struggle to produce factually correct descriptions without user-written guidelines based on automatic and human evaluation. Common error patterns include guideline misalignment, incorrect variables, missing code details, and faulty reasoning. Ablations show that implementing guidelines substantially improves performance. Our downstream application examining Human-AI collaboration in the documentation for code-table demonstrated enhanced usability and effectiveness.

NBDESCRIB: A Dataset for Text Description Generation from Tables and Code in Jupyter Notebooks with Guidelines

Recent advances in video-text retrieval (VTR) have largely relied on supervised learning and fine-tuning. In this paper, we introduce , a novel zero-shot VTR framework that leverages off-the-shelf video captioners, large language models (LLMs), and text retrieval methods—entirely without additional training or annotated data. Due to the limited power of captioning methods, the captions often miss important content in the video, resulting in unsatisfactory retrieval performance. To translate more information into video captions, we first generates initial captions for videos, then enhances them using a relevance-boosted captioning strategy powered by LLMs, enriching video descriptions with salient details. To further emphasize key content, we propose structural information extraction, organizing visual elements such as objects, events, and attributes into structured templates, further boosting the retrieval performance. Benefiting from the enriched captions and structuralized information, extensive experiments on several video-text retrieval benchmarks demonstrate the superiority of over existing fine-tuned and pretraining methods without any data. They also show that the enriched captions capture key details from the video with minimal noise. Code and data will be released to facilitate future research.

ELIOT: Zero-Shot Video-Text Retrieval through Relevance-Boosted Captioning and Structural Information Extraction

Commonsense inference and domain-specific expertise are crucial for understanding and responding to emotional, cognitive, and topic-specific cues in counseling conversations with crime victims. However, these key evidences are often dispersed across multiple utterances, making it difficult to capture through single-hop reasoning. To address this, we propose MENDER, a novel Multi-hop commonsensE and domaiN-specific Chain-of-Thought (CoT) reasoning framework for knowleDge-grounded empathEtic Response generation in counseling dialogues. MENDER leverages large language models (LLMs) to integrate commonsense and domain knowledge via multi-hop reasoning over the dialogue context. It employs two specialized reasoning chains, viz. Commonsense Knowledge-driven CoT and Domain Knowledge-driven CoT rationales, which extract and aggregate dispersed emotional, cognitive, and topical evidences to generate knowledge-grounded empathetic counseling responses. Experimental evaluations on counseling dialogue dataset, POEM validate MENDER’s efficacy in generating coherent, empathetic, knowledge-grounded responses.

MENDER: Multi-hop Commonsense and Domain-specific CoT Reasoning for Knowledge-grounded Empathetic Counseling of Crime Victims

Large language models (LLMs) have gained popularity in recent years for their utility in various applications. However, they are sensitive to non-semantic changes in prompt formats, where small changes in the prompt format can lead to significant performance fluctuations. In the literature, this problem is commonly referred to as prompt brittleness. Previous research on prompt engineering has focused mainly on developing techniques for identifying the optimal prompt for specific tasks. Some studies have also explored the issue of prompt brittleness and proposed methods to quantify performance variations; however, no simple solution has been found to address this challenge. We propose Mixture of Formats (MOF), a simple and efficient technique for addressing prompt brittleness in LLMs by diversifying the styles used in the prompt few-shot examples. MOF was inspired by computer vision techniques that utilize diverse style datasets to prevent models from associating specific styles with the target variable. Empirical results show that our proposed technique reduces style-induced prompt brittleness in various LLMs while also enhancing overall performance across prompt variations and different datasets.

Downloads

Next from NAACL 2025

HyPA-RAG: A Hybrid Parameter Adaptive Retrieval-Augmented Generation System for AI Legal and Policy Applications

Similar lecture

PTQ1.61: Push the Real Limit of Extremely Low-Bit Post-Training Quantization Methods for Large Language Models