China

Retrieval-augmented generation (RAG) enhances the outputs of language models by integrating relevant information retrieved from external knowledge sources. However, when the retrieval process involves private data, RAG systems may face severe privacy risks, potentially leading to the leakage of sensitive information. To address this issue, we propose using synthetic data as a privacy-preserving alternative for the retrieval data. We propose SAGE, a novel two-stage synthetic data generation paradigm. In the stage-1, we employ an attribute-based extraction and generation approach to preserve key contextual information from the original data. In the stage-2, we further enhance the privacy properties of the synthetic data through an agent-based iterative refinement process. Extensive experiments demonstrate that using our synthetic data as the retrieval context achieves comparable performance to using the original data while substantially reducing privacy risks. Our work takes the first step towards investigating the possibility of generating high-utility and privacy-preserving synthetic data for RAG, opening up new opportunities for the safe application of RAG systems in various domains.

EMNLP 2025

Mitigating the Privacy Issues in Retrieval-Augmented Generation (RAG) via Pure Synthetic Data

synthetic data

privacy

### Welcome!
The 62nd Annual Meeting of the Association for Computational Linguistics (ACL 2024) will take place in Bangkok, Thailand from August 11th to 16th, 2024. Our Virtual Poster Sessions will take place online Thursday, August 22, 2024.

You are required to register for this event. **Please register [here](https://2024.aclweb.org/registration). **

If you have already registered, please check your inbox for an email from Underline granting you access to ACL 2024 content.

Please register!

ACL 2024

The 62nd Annual Meeting of the Association for Computational Linguistics (ACL 2024) will take place in Bangkok, Thailand from August 11th to 16th, 2024. More information will be announced soon.

poster

## 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*

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

Auditing autoregressive Large Language Models (LLMs) for disparities is often impeded by high token costs and limited precision. We introduce Token-Focused Disparity Probing (TFDP), a novel methodology overcoming these challenges by adapting single-token masked prediction to autoregressive architectures via targeted token querying. Disparities between minimally contrastive sentence pairs are quantified through a multi-scale semantic alignment score that integrates sentence, local-context, and token embeddings with adaptive weighting. We propose three disparity metrics: Preference Score (mathcalPS), Prediction Set Divergence (mathcalPSD), and Weighted Final Score (mathcalWFS), for comprehensive assessment. Evaluated on our customized Proverbs Disparity Dataset (PDD) with controlled attribute toggles (e.g., gender bias, misinformation susceptibility), TFDP precisely detects disparities while achieving up to 42 times fewer output tokens than minimal n-token continuations, offering a scalable tool for responsible LLM evaluation.

TFDP: Token-Efficient Disparity Audits for Autoregressive LLMs via Single-Token Masked Evaluation

As the adoption of language models advances, so does the need to better represent individual users to the model. Are there aspects of an individual's belief system that a language model can utilize for improved alignment? Following prior research, we investigate this question in the domain of opinion prediction by developing PrimeX, a dataset of public opinion survey data from over 800 US residents with two additional sources of belief information: written explanations from the respondents for why they hold specific opinions, and the Primal World Belief survey for assessing respondent worldview. We provide an extensive initial analysis of our data and show the value of belief explanations and worldview for personalizing language models. Our results demonstrate how the additional belief information in PrimeX can benefit both the NLP and psychological research communities and open up avenues for further study.

PrimeX: A Dataset of Worldview, Opinion, and Explanation

Multimodal Retrieval-Augmented Generation (MRAG) systems enhance LMMs by integrating external multimodal databases, but introduce unexplored privacy vulnerabilities. While text-based RAG privacy risks have been studied, multimodal data presents unique challenges. We provide the first systematic analysis of MRAG privacy vulnerabilities across vision-language and speech-language modalities. Using a novel compositional structured prompt attack in a black-box setting, we demonstrate how attackers can extract private information by manipulating queries. Our experiments reveal that LMMs can both directly generate outputs resembling retrieved content and produce descriptions that indirectly expose sensitive information, highlighting the urgent need for robust privacy-preserving MRAG techniques.

Beyond Text: Unveiling Privacy Vulnerabilities in Multi-modal Retrieval-Augmented Generation

With the rise of Speech Large Language Models (SpeechLLMs), two dominant approaches have emerged for speech processing: discrete tokens and continuous features. Each approach has demonstrated strong capabilities in audio-related processing tasks. However, the performance gap between these two paradigms has not been thoroughly explored. To address this gap, we present a fair comparison of self-supervised learning (SSL)-based discrete and continuous features under the same experimental settings. We evaluate their performance across six spoken language understanding-related tasks using both small and large-scale LLMs (Qwen1.5-0.5B and Llama3.1-8B). We further conduct in-depth analyses, including efficient comparison, SSL layer analysis, LLM layer analysis, and robustness comparison. Our findings reveal that continuous features generally outperform discrete tokens in various tasks. Each speech processing method exhibits distinct characteristics and patterns in how it learns and processes speech information. We hope our results will provide valuable insights to advance spoken language understanding in SpeechLLMs.

Speech Discrete Tokens or Continuous Features? A Comparative Analysis for Spoken Language Understanding in SpeechLLMs

Large language models (LLMs) are increasingly used as evaluators in natural language generation tasks, offering advantages in scalability and interpretability over traditional evaluation methods. However, existing LLM-based evaluations often suffer from biases and misalignment, particularly in domain-specific tasks, due to limited functional understanding and knowledge gaps. To address these challenges, we first investigate the relationship between an LLM-based evaluator’s familiarity with the target task and its evaluation performance. We then introduce the Co-Eval framework, which leverages a criteria planner model and optimized machine metrics to enhance the scalability and fairness of LLM-based evaluation. Experimental results on both general and domain-specific tasks demonstrate that Co-Eval reduces biases, achieving up to a 0.4903 reduction in self-preference bias, and improves alignment with human preferences, with gains of up to 0.324 in Spearman correlation.

Co-Eval: Augmenting LLM-based Evaluation with Machine Metrics

Large Language Models (LLMs) have shown promise in simulating human social behavior and interactions, yet lack large-scale, systematically constructed benchmarks for evaluating their alignment with real-world social tendencies. To address this gap, we introduce \textbf{SocioBench}, a comprehensive, cross-national, and cross-topic sociological benchmark derived from the \textit{International Social Survey Programme }(ISSP). It encompasses over 480,000 real respondent records from more than 30 countries, spanning 10 sociological issues, and over 40 detailed demographic attributes. Our experiments reveal notable variations in the accuracy and systematic biases in the responses of current mainstream open-source LLMs when simulating social attitudes across different topics and demographic groups.

SocioBench: Modeling Human Behavior in Sociological Surveys with Large Language Models

Vision-language models demand watermarking solutions that protect intellectual property without compromising multimodal coherence. Existing text watermarking methods disrupt visual-textual alignment through biased token selection and static strategies, leaving semantic-critical concepts vulnerable. We propose VLA-Mark, a vision-aligned framework that embeds detectable watermarks while preserving semantic fidelity through cross-modal coordination. Our approach integrates multiscale visual-textual alignment metrics, combining localized patch affinity, global semantic coherence, and contextual attention patterns, to guide watermark injection without model retraining. An entropy-sensitive mechanism dynamically balances watermark strength and semantic preservation, prioritizing visual grounding during low-uncertainty generation phases. Experiments show 7.4% lower PPL and 26.6% higher BLEU than conventional methods, with near-perfect detection (98.8% AUC). The framework demonstrates 96.1% attack resilience against attacks such as paraphrasing and synonym substitution, while maintaining text-visual consistency, establishing new standards for quality-preserving multimodal watermarking.

VLA-Mark: A cross modal watermark for large vision-language alignment models

Dehumanization, i.e., denying human qualities to individuals or groups, is a particularly harmful form of hate speech that can normalize violence against marginalized communities. Despite advances in NLP for detecting general hate speech, approaches to identifying dehumanizing language remain limited due to scarce annotated data and the subtle nature of such expressions. In this work, we systematically evaluate four state-of-the-art large language models (LLMs) — Claude, GPT, Mistral, and Qwen — for dehumanization detection. Our results show that only one model—Claude—achieves strong performance (over 80% F₁) under an optimized configuration, while others, despite their capabilities, perform only moderately. Performance drops further when distinguishing dehumanization from related hate types such as derogation. We also identify systematic disparities across target groups: models tend to over-predict dehumanization for some identities (e.g., Gay men), while under-identifying it for others (e.g., Refugees). These findings motivate the need for systematic, group-level evaluation when applying pretrained language models to dehumanization detection tasks.

Beyond Hate Speech: NLP’s Challenges and Opportunities in Uncovering Dehumanizing Language

Trustworthiness in healthcare question-answering (QA) systems is important for ensuring patient safety, clinical effectiveness, and user confidence. As large language models (LLMs) become increasingly integrated into medical settings, the reliability of their responses directly influences clinical decision-making and patient outcomes. However, achieving comprehensive trustworthiness in medical QA poses significant challenges due to the inherent complexity of healthcare data, the critical nature of clinical scenarios, and the multifaceted dimensions of trustworthy AI. In this survey, we systematically examine six key dimensions of trustworthiness in medical QA, i.e., Factuality, Robustness, Fairness, Safety, Explainability, and Calibration. We review how each dimension is evaluated in existing LLM‐based medical QA systems. We compile and compare major benchmarks designed to assess these dimensions and analyze evaluation‐guided techniques that drive model improvements, such as retrieval‐augmented grounding, adversarial fine‐tuning, and safety alignment. Finally, we identify open challenges—such as scalable expert evaluation, integrated multi‐dimensional metrics, and real‐world deployment studies—and propose future research directions to advance the safe, reliable, and transparent deployment of LLM‐powered medical QA.

Trustworthy Medical Question Answering: An Evaluation-Centric Survey

Large language model (LLM)-driven multi-agent systems (MAS) are transforming how humans and AIs collaboratively generate ideas and artifacts. While existing surveys provide comprehensive overviews of MAS infrastructures, they largely overlook the dimension of \emph{creativity}, including how novel outputs are generated and evaluated, how creativity informs agent personas, and how creative workflows are coordinated. This is the first survey dedicated to creativity in MAS. We focus on text and image generation tasks, and present: (1) a taxonomy of agent proactivity and persona design; (2) an overview of generation techniques, including divergent exploration, iterative refinement, and collaborative synthesis, as well as relevant datasets and evaluation metrics; and (3) a discussion of key challenges, such as inconsistent evaluation standards, insufficient bias mitigation, coordination conflicts, and the lack of unified benchmarks. This survey offers a structured framework and roadmap for advancing the development, evaluation, and standardization of creative MAS.

Downloads

Next from EMNLP 2025

TFDP: Token-Efficient Disparity Audits for Autoregressive LLMs via Single-Token Masked Evaluation

Similar lecture

The Good and The Bad: Exploring Privacy Issues in Retrieval-Augmented Generation (RAG)