Austria

Reasoning is an intricate process that transcends both language and vision; yet, despite its inherently modality-agnostic nature, develop-
ing effective multilingual and multimodal reasoning capabilities remains a substantial challenge for Multimodal Large Language Models (MLLMs). They struggle to activate complex reasoning behaviours, delivering step-wise explanation, questioning and reflection, particularly in multilingual settings where high-quality supervision across languages is lacking. Recent works have introduced eclectic strategies to enhance MLLMs’ reasoning; however, they remain related to a single language.
To make MLLMs’ reasoning capabilities aligned among languages and improve modality performances, we propose R2-MultiOmnia, a modular approach that instructs the models to abstract key elements of the reasoning process and then refine reasoning trajectories via self-correction. Specifically, we instruct the models producing multimodal synthetic resources by bridging modalities and then self-improving their capabilities. To stabilise learning and the reasoning processes structure, we propose Curriculum Learning Reasoning Stabilisation with structured output rewards to gradually refine the models’ capabilities to learn and deliver robust reasoning processes. Experiments show that R2-MultiOmnia improves multimodal reasoning, gets aligned performances among the languages approaching strong models.

ACL 2025

x-SAL: Leading Symbolic Reasoning across Languages via Cross-lingual Symbolic-Aided Language Model

multilingual reasoning

## Welcome!
"I am excited to welcome you to this year’s edition of the Conference on Empirical Methods in Natural Language Processing! Importantly, it marks the 30th edition of EMNLP. With over 8,000 submissions, more than 3,000 accepted papers, and thousands of attendees, we have come a long way from that first
workshop, which had 14 accepted papers. As the field looks ahead, Suzhou is the fitting location for celebrating this milestone: rooted in a long literary tradition, yet modern and forward-looking, and home to a large share of the NLP community." 

*Message from the General Chair, Dirk Hovy*

EMNLP 2025

Celebrate 30 Years of EMNLP! 
EMNLP 2025 will be held in Suzhou, China from November 5th to November 9th, 2025.

poster

### 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**. 

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

Large Language Models (LLMs) have achieved remarkable success in Natural Language Processing (NLP), yet their cross-lingual consistency remains a significant challenge. This paper introduces a novel methodology for efficiently identifying inherent cross-lingual weaknesses in LLMs. Our approach leverages beam search and LLM-based simulation to generate bilingual question pairs that expose performance discrepancies between English and target languages. We construct a new dataset of over 6,000 bilingual pairs across 16 languages using this methodology, demonstrating its effectiveness in revealing weaknesses even in state-of-the-art models. The extensive experiments demonstrate that our method precisely and cost-effectively pinpoints cross-lingual weaknesses, consistently revealing over 50% accuracy drops in target languages across a wide range of models. Moreover, further experiments investigate the relationship between linguistic similarity and cross-lingual weaknesses, revealing that linguistically related languages share similar performance patterns and benefit from targeted post-training. Code is available at https://github.com/xzx34/Cross-Lingual-Pitfalls.

Cross-Lingual Pitfalls: Automatic Probing Cross-Lingual Weakness of Multilingual Large Language Models

The paper explores the performance of LLMs in the context of multi-dimensional analytic writing assessments, i.e. their ability to provide both scores and comments based on multiple assessment criteria. Using a corpus of literature reviews written by L2 graduate students and assessed by human experts against 9 analytic criteria, we prompt several popular LLMs to perform the same task under various conditions. To evaluate the quality of feedback comments, we apply a novel feedback comment quality evaluation framework. This framework is interpretable, cost-efficient, scalable, and reproducible, compared to existing methods that rely on manual judgments. We find that LLMs can generate reasonably good and generally reliable multi-dimensional analytic assessments. We release our corpus and code for reproducibility.

LLMs can Perform Multi-Dimensional Analytic Writing Assessments: A Case Study of L2 Graduate-Level Academic English Writing

Despite progress in comment-aware multimodal and multilingual summarization for English and Chinese, research in Indian languages remains limited. This study addresses this gap by introducing COSMMIC, a pioneering comment-sensitive multimodal, multilingual dataset featuring nine major Indian languages. COSMMIC comprises 4,959 article-image pairs and 24,484 reader comments, with ground-truth summaries available in all included languages. Our approach enhances summaries by integrating reader insights and feedback. We explore summarization and headline generation across four configurations: (1) using article text alone, (2) incorporating user comments, (3) utilizing images, and (4) combining text, comments, and images. To assess the dataset’s effectiveness, we employ state-of-the-art language models such as LLama3 and GPT-4. We conduct a comprehensive study to evaluate different component combinations, including identifying supportive comments, filtering out noise using a dedicated comment classifier using IndicBERT, and extracting valuable insights from images with a multilingual CLIP-based classifier. This helps determine the most effective configurations for natural language generation (NLG) tasks. Unlike many existing datasets that are either text-only or lack user comments in multimodal settings, COSMMIC uniquely integrates text, images, and user feedback. This holistic approach bridges gaps in Indian language resources, advancing NLP research and fostering inclusivity.

COSMMIC: Comment-Sensitive Multimodal Multilingual Indian Corpus for Summarization and Headline Generation

Predicting semantic textual similarity (STS) is a complex and ongoing challenge in natural language processing (NLP). Over the years, researchers have developed a variety of supervised and unsupervised approaches to calculate STS automatically. Additionally, various benchmarks, which include STS datasets, have been established to consistently evaluate and compare these STS methods. However, they largely focus on high-resource languages, mixed with datasets annotated focusing on relatedness instead of similarity and containing automatically translated instances. Therefore, no dedicated benchmark for multilingual STS exists. To solve this gap, we introduce the Multilingual Semantic Textual Similarity Benchmark (MUSTS), which spans 13 languages, including low-resource languages. By evaluating more than 25 models on MUSTS, we establish the most comprehensive benchmark of multilingual STS methods. Our findings confirm that STS remains a challenging task, particularly for low-resource languages.

MUSTS: MUltilingual Semantic Textual Similarity Benchmark

Large Language Models (LLMs) are often used as automated judges to evaluate text, but their effectiveness can be hindered by various unintentional biases. We propose using linear classifying probes, trained by leveraging differences between contrasting pairs of prompts, to directly access LLMs' latent knowledge and extract more accurate preferences. Through extensive experiments using models of varying size from four different families and six diverse datasets assessing text quality evaluation and common sense reasoning, we demonstrate that both supervised and unsupervised probing approaches consistently outperform traditional generation-based judgement while maintaining similar computational costs. These probes generalise under domain shifts and can even outperform finetuned evaluators with the same training data size. Our results suggest linear probing offers an accurate, robust and computationally efficient approach for LLM-as-judge tasks while providing interpretable insights into how models encode judgement-relevant knowledge. Our data and code will be openly released in the future.

Improving Preference Extraction In LLMs By Identifying Latent Knowledge Through Classifying Probes

Pretrained language models have significantly advanced performance across various natural language processing tasks. However, adversarial attacks continue to pose a critical challenge to system built using these models, as they can be exploited with carefully crafted adversarial texts. Inspired by the ability of diffusion models to predict and reduce noise in computer vision, we propose a novel and flexible adversarial defense method for language classification tasks, DiffuseDef, which incorporates a diffusion layer as a denoiser between the encoder and the classifier. The diffusion layer is trained on top of the existing classifier, ensuring seamless integration with any model in a plug-and-play manner. During inference, the adversarial hidden state is first combined with sampled noise, then denoised iteratively and finally ensembled to produce a robust text representation. By integrating adversarial training, denoising, and ensembling techniques, we show that DiffuseDef improves over existing adversarial defense methods and achieves state-of-the-art performance against common black-box and white-box adversarial attacks.

DiffuseDef: Improved Robustness to Adversarial Attacks via Iterative Denoising

Skewness in imbalanced datasets affects Automatic Text Classification (ATC), leading to classifier bias toward the majority classes. This work examines undersampling methods to mitigate such bias in Small and Large Language Model (SLMs and LLMs) classifiers. Based on the limitations found in existing solutions, we propose two novel undersampling methods inspired by state-of-the-art Instance Selection techniques, relying on calibrated confidences and semantic difficulty estimates. We compare them against 19 baselines across 13 datasets, evaluating: (i) effectiveness, (ii) class imbalance bias, (iii) efficiency, (iv) scalability, and (v) consistency. Results show our methods uniquely reduce classifier bias (up to 56%) across all datasets without effectiveness loss while improving efficiency (1.6x speedup), scalability and reducing carbon emissions (up to 50%).

Instance-Selection-Inspired Undersampling Strategies for Bias Reduction in Small and Large Language Models for Binary Text Classification

Parameter-efficient fine-tuning (PEFT) addresses the memory footprint issue of full fine-tuning by modifying only a subset of model parameters. However, on datasets exhibiting spurious correlations, we observed that PEFT slows down the model's convergence on unbiased examples, while the convergence on biased examples remains fast. This leads to the model's overfitting on biased examples, causing significant performance degradation in out-of-distribution (OOD) scenarios. Traditional debiasing methods mitigate this issue by emphasizing unbiased examples during training but often come at the cost of in-distribution (ID) performance drops. To address this trade-off issue, we propose a curriculum debiasing framework that presents examples in a \textit{biased-to-unbiased} order. Our framework initially limits the model’s exposure to unbiased examples, which are harder to learn, allowing it to first establish a foundation on easier-to-converge biased examples. As training progresses, we gradually increase the proportion of unbiased examples in the training set, guiding the model away from reliance on spurious correlations. Compared to the original PEFT methods, our method accelerates convergence on unbiased examples by approximately twofold and improves ID and OOD performance by 1.2% and 8.0%, respectively.

Curriculum Debiasing: Toward Robust Parameter-Efficient Fine-Tuning Against Dataset Biases

The large language model (LLM)-as-judge paradigm has been used to meet the demand for a cheap, reliable, and fast evaluation of model outputs during AI system development and post-deployment monitoring. While judge models—LLMs finetuned to specialize in assessing and critiquing model outputs—have been touted as general purpose evaluators, they are typically evaluated only on non-contextual scenarios, such as instruction following. The omission of contextual settings—those where external information is used as context to generate an output—is surprising given the increasing prevalence of retrieval-augmented generation (RAG) and summarization use cases. Contextual assessment is uniquely challenging, as evaluation often depends on practitioner priorities, leading to conditional evaluation criteria (e.g., comparing responses based on factuality and then considering completeness if they are equally factual). To address the gap, we propose ContextualJudgeBench, a judge benchmark with 2,000 challenging response pairs across eight splits inspired by real-world contextual evaluation scenarios. We build our benchmark with a multi-pronged data construction pipeline that leverages both existing human annotations and model-based perturbations. Our comprehensive study across 11 judge models and 7 general purpose models, reveals that the contextual information and assessment criteria present a significant challenge to even state-of-the-art models. For example, o1, the best-performing model, barely reaches 55% consistent accuracy.

Does Context Matter? ContextualJudgeBench for Evaluating LLM-based Judges in Contextual Settings

Large Language Models (LLMs) demonstrate exceptional performance across diverse tasks by leveraging pre-trained (i.e., parametric) and external (i.e., contextual) knowledge. While substantial efforts have been made to enhance the utilization of both forms of knowledge, situations in which models lack relevant information remain underexplored. To investigate this challenge, we first present a controlled testbed featuring four distinct knowledge access scenarios, including the aforementioned edge case, revealing that conventional LLM usage exhibits insufficient robustness in handling all instances. Addressing this limitation, we propose Contrastive Decoding with Abstention (CDA), a novel training-free decoding method that allows LLMs to generate responses when relevant knowledge is available and to abstain otherwise. CDA estimates the relevance of both knowledge sources for a given input, adaptively deciding which type of information to prioritize and which to exclude. Through extensive experiments, we demonstrate that CDA can effectively perform accurate generation and abstention simultaneously, enhancing reliability and preserving user trust.

Downloads

Next from ACL 2025

Cross-Lingual Pitfalls: Automatic Probing Cross-Lingual Weakness of Multilingual Large Language Models

Similar lecture

A Survey of Multilingual Reasoning in Language Models