Austria

Recent studies on interpreting the hidden states of speech models have shown their ability to capture speaker-specific features, including gender. Does this finding also hold for speech translation (ST) models? If so, what are the implications for the speaker&#39;s gender assignment in translation? We address these questions from an interpretability perspective, using probing methods to assess gender encoding across diverse ST models. Results on three language directions (English $\rightarrow$ French/Italian/Spanish) indicate that while traditional encoder-decoder models capture gender information, newer architectures---integrating a speech encoder with a machine translation system via adapters---do not. We also demonstrate that low gender encoding capabilities result in systems&#39; tendency toward a masculine default, a translation bias that is more pronounced in newer architectures.

ACL 2025

Different Speech Translation Models Encode and Translate Speaker Gender Differently

gender bias

interpretability

speech translation

Recent studies on interpreting the hidden states of speech models have shown their ability to capture speaker-specific features, including gender. Does this finding also hold for speech translation (ST) models? If so, what are the implications for the speaker's gender assignment in translation? We address these questions from an interpretability perspective, using probing methods to assess gender encoding across diverse ST models. Results on three language directions (English $\rightarrow$ French/Italian/Spanish) indicate that while traditional encoder-decoder models capture gender information, newer architectures---integrating a speech encoder with a machine translation system via adapters---do not. We also demonstrate that low gender encoding capabilities result in systems' tendency toward a masculine default, a translation bias that is more pronounced in newer architectures.

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 are required to register.
Please complete your registration to continue.

We recommend reading [**the registration details**](https://2025.aclweb.org/registration/) first.

Registration Required

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

Inference methods play an important role in eliciting the performance of large language models (LLMs). Currently, LLMs use inference methods utilizing generated multiple samples, which can be derived from Minimum Bayes Risk (MBR) Decoding. Previous studies have conducted empirical analyses to clarify the improvements in generation performance achieved by MBR decoding and have reported various observations. However, the theoretical underpinnings of these findings remain uncertain. To address this, we offer a new theoretical interpretation of MBR decoding from the perspective of bias–diversity decomposition. In this interpretation, the error in the quality estimation of hypotheses by MBR decoding is decomposed into two main factors: bias, which considers the closeness between the utility function and human evaluation, and diversity, which represents the variability in the quality estimation of the utility function. The theoretical analysis reveals the difficulty of simultaneously improving bias and diversity, confirming the validity of enhancing MBR decoding performance by increasing diversity. Furthermore, we reveal that diversity can explain one aspect of inference scaling laws that describe performance improvement by increasing sample size. Moreover, experiments across multiple NLP tasks yielded results consistent with these theoretical characteristics. Our code is available at https://github.com/naist-nlp/mbr-bias-diversity.

Comprehensive Analysis of Minimum Bayes Risk Decoding through Bias and Diversity Decomposition

Large language models (LLMs) deliver impressive results but face challenges from increasing model sizes and computational costs. Structured pruning reduces model size and speeds up inference but often causes uneven degradation across domains, leading to biased performance. To address this, we propose *DRPruning*, a method that dynamically adjusts the data distribution during training to restore balanced performance across heterogeneous and multi-tasking data. Experiments in monolingual and multilingual settings show that DRPruning surpasses similarly sized models in both pruning and continued pretraining over perplexity, downstream tasks, and instruction tuning. Further analysis demonstrates the robustness of DRPruning towards various domains and distribution shifts. Furthermore, DRPruning can determine optimal reference losses and data ratios automatically, suggesting potential for broader applications. Code and scripts are available at https://github.com/hexuandeng/DRPruning.

DRPruning: Efficient Large Language Model Pruning through Distributionally Robust Optimization

With a combination of quantitative experiments, human judgments, and qualitative analyses, we evaluate the quantity and quality of African American Language (AAL) representation in 12 predominantly English, open-source pretraining corpora. We specifically focus on the sources, variation, and naturalness of included AAL texts representing the AAL speaking community. We find that AAL is underrepresented in all evaluated pretraining corpora compared to US demographics, constituting as few as 0.007% and at most 0.18% of documents. We also find that more than 25% of AAL texts in C4 may be perceived as inappropriate for LLMs to generate and to reinforce harmful stereotypes. Finally, we find that most automated filters are more likely to conserve White Mainstream English (WME) texts over AAL in pretraining corpora.

Data Caricatures: On the Representation of African American Language in Pretraining Corpora

Developing robust automatic speech recognition (ASR) systems for Arabic requires effective strategies to manage its diversity. Existing ASR systems mainly cover the modern standard Arabic (MSA) variety and few high-resource dialects, but fall short in coverage and generalization across the multitude of spoken variants. Code-switching with English and French is also common in different regions of the Arab world, which challenges the performance of monolingual Arabic models. In this work, we introduce a suite of ASR models optimized to effectively recognize multiple variants of spoken Arabic, including MSA, various dialects, and code-switching. We provide open-source pre-trained models that cover data from 17 Arabic-speaking countries, and fine-tuned MSA and dialectal ASR models that include at least 11 variants, as well as multi-lingual ASR models covering embedded languages in code-switched utterances. We evaluate ASR performance across these spoken varieties and demonstrate both coverage and performance gains compared to prior models.

Dialectal Coverage And Generalization in Arabic Speech Recognition

Iterative non-autoregressive (NAR) models share a spirit of mixed autoregressive (AR) and fully NAR models, seeking a balance between generation quality and inference efficiency. These models have recently demonstrated impressive performance in varied generation tasks, surpassing the autoregressive Transformer. However, they also face several challenges that impede further development. In this work, we target building more efficient and competitive iterative NAR models. Firstly, we produce two simple metrics to identify the potential problems existing in current refinement processes, and look back on the various iterative NAR models to find the key factors for realizing our purpose. Subsequently, based on the analyses of the limitations of previous inference algorithms, we propose a simple yet effective strategy to conduct efficient refinements without performance declines. Experiments on five widely used datasets show that our final models set the new state-of-the-art performance compared to all previous NAR models, even with fewer decoding steps, and outperform AR Transformer by around one BLEU on average. Our codes and models are available on Github.

An Empirical Study of Iterative Refinements for Non-autoregressive Translation

We introduce GrammaMT, a grammatically-aware prompting approach for machine translation that uses Interlinear Glossed Text (IGT), a common form of linguistic description providing morphological and lexical annotations for source sentences. GrammaMT proposes three prompting strategies: gloss-shot, chain-gloss and model-gloss. All are training-free, requiring only a few examples that involve minimal effort to collect, and making them well-suited for low-resource setups. Experiments show that GrammaMT enhances translation performance on open-source instruction-tuned LLMs for various low- to high-resource languages across three benchmarks: (1) the largest IGT corpus, (2) the challenging 2023 SIGMORPHON Shared Task data over endangered languages, and (3) even in an out-of-domain setting with FLORES. Moreover, ablation studies reveal that leveraging gloss resources could substantially boost MT performance (by over 17 BLEU points) if LLMs accurately generate or access input sentence glosses.

GrammaMT: Improving Machine Translation with Grammar-Informed In-Context Learning

Chart generation aims to generate code to produce charts satisfying the desired visual properties, e.g., texts, layout, color, and type. It has great potential to empower the automatic professional report generation in financial analysis, research presentation, education, and healthcare. In this work, we build a vision-language model (VLM) based multi-agent framework for effective automatic chart generation. Generating high-quality charts requires both strong visual design skills and precise coding capabilities that embed the desired visual properties into code. Such a complex multi-modal reasoning process is difficult for direct prompting of VLMs. To resolve these challenges, we propose METAL, a multi-agent framework that decomposes the task of chart generation into the iterative collaboration among specialized agents. METAL achieves a 5.2\% improvement in the F1 score over the current best result in the chart generation task. Additionally, METAL improves chart generation performance by 11.33% over Direct Prompting with LLaMA-3.2-11B.
Furthermore, the METAL framework exhibits the phenomenon of test-time scaling: its performance increases monotonically as the logarithm of computational budget grows from 512 to 8192 tokens.

METAL: A Multi-Agent Framework for Chart Generation with Test-Time Scaling

The widespread success of LLMs on NLP benchmarks has been accompanied by concerns that LLMs function primarily as stochastic parrots that reproduce texts similar to what they saw during pre-training, often erroneously. But what is the nature of their errors, and do these errors exhibit any regularities? In this work, we examine irrelevant context hallucinations, in which models integrate misleading contextual cues into their predictions. Through behavioral analysis, we show that these errors result from a structured yet flawed mechanism that we term _class-based (mis)generalization_, in which models combine abstract class cues with features extracted from the query or context to derive answers. Furthermore, mechanistic interpretability experiments on Llama-3, Mistral, and Pythia across 39 factual recall relation types reveal that this behavior is reflected in the model's internal computations: (i) abstract class representations are constructed in lower layers before being refined into specific answers in higher layers, (ii) feature selection is governed by two competing circuits --- one prioritizing direct query-based reasoning, the other incorporating contextual cues --- whose relative influences determine the final output. Our findings provide a more nuanced perspective on the stochastic parrot argument: through form-based training, LLMs can exhibit generalization leveraging abstractions, albeit in unreliable ways based on contextual cues — what we term _stochastic chameleons_.

Stochastic Chameleons: Irrelevant Context Hallucinations Reveal Class-Based (Mis)Generalization in LLMs

Machine unlearning (MU) has gained significant attention as a means to remove the influence of specific data from a trained model without requiring full retraining. While progress has been made in unimodal domains like text and image classification, unlearning in multimodal models remains relatively under-explored. In this work, we address the unique challenges of unlearning in CLIP, a prominent multimodal model that aligns visual and textual representations. We introduce CLIPErase, a novel approach that disentangles and selectively forgets both visual and textual associations, ensuring that unlearning does not compromise model performance.
CLIPErase consists of three key modules: a Forgetting Module that disrupts the associations in the forget set, a Retention Module that preserves performance on the retain set, and a Consistency Module that maintains consistency with the original model. Extensive experiments on CIFAR-100, Flickr30K, and Conceptual 12M across five CLIP downstream tasks, as well as an evaluation on diffusion models, demonstrate that CLIPErase effectively removes designated associations from multimodal samples in downstream tasks, while preserving the model’s performance on the retain set after unlearning.

CLIPErase: Efficient Unlearning of Visual-Textual Associations in CLIP

Entity grounding, a crucial task in constructing multimodal knowledge graphs, aims to align entities from knowledge graphs with their corresponding images. Unlike conventional visual grounding tasks that use referring expressions (REs) as inputs, entity grounding relies solely on entity names and types, presenting a significant challenge. To address this, we introduce a novel **I**terative **S**elf-**R**efinement (**ISR**) scheme to enhance the multimodal large language model's capability to generate high quality REs for the given entities as explicit contextual clues. This training scheme, inspired by human learning dynamics and human annotation processes, enables the MLLM to iteratively generate and refine REs by learning from successes and failures, guided by outcome rewards from a visual grounding model. This iterative cycle of self-refinement avoids overfitting to fixed annotations and fosters continued improvement in referring expression generation. Extensive experiments demonstrate that our methods surpasses other methods in entity grounding, highlighting its effectiveness, robustness and potential for broader applications.

Downloads

Next from ACL 2025

Comprehensive Analysis of Minimum Bayes Risk Decoding through Bias and Diversity Decomposition

Stay up to date with the latest Underline news!

PRESENTATIONS

CONFERENCES

COMPANY

RESOURCES