China

Language models (LMs) hallucinate. We inquire: Can we detect and mitigate hallucinations before they happen? This work answers this research question in the positive, by showing that the internal representations of LMs provide rich signals that can be used for this purpose. We introduce FactCheckmate, which preemptively detects hallucinations by learning a classifier that predicts whether the LM will hallucinate, based on the model&#39;s hidden states produced over the inputs, before decoding begins. If a hallucination is detected, FactCheckmate then intervenes by adjusting the LM&#39;s hidden states such that the model will produce more factual outputs. FactCheckmate provides fresh insights that the inner workings of LMs can be revealed by their hidden states. Practically, both its detection and mitigation models are lightweight, adding little inference overhead; FactCheckmate proves a more efficient approach for mitigating hallucinations compared to many post-hoc alternatives. We evaluate FactCheckmate over LMs of different scales and model families (including Llama, Mistral, Qwen and Gemma),across a variety of QA datasets from different domains. Our results demonstrate the effectiveness of FactCheckmate, achieving over 70% preemptive detection accuracy. On average, outputs generated by LMs with intervention are 34.4% more factual compared to those without.

EMNLP 2025

FACTCHECKMATE: Preemptively Detecting and Mitigating Hallucinations in LMs

hallucination mitigation

hallucination detection

large language models

factuality

It is my great pleasure and honor to welcome you to the 64th Annual Meeting of the Association for Computational Linguistics (ACL 2026), held in San Diego, California, USA, July 2–7, 2026. ACL 2026 continues our fields 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.

It is exciting to see the many developments in the field reflected at our conference, such as the advancement of large language models (born out of our community) into many application directions. This is not without challenges, partly due to unsolved problems, and partly due to misuse of the technology. All this is reflected well at ACL 2026. The conference also reflects that the field is not only about advancing technology, but also the science of fundamentally understanding language and how it is use by many different speaker communities. 
[Full Message from General Chair Philipp Koehn](https://drive.google.com/file/d/1QiJFOmMjvDgsDo3i1XzlQnImxe3RU_Cy/view?usp=sharing) 
<html><button style="display: inline-flex; align-items: center; justify-content: center; white-space: nowrap; border-radius: 9999px; font-weight: bold; background: #7c3aed; color: white; font-family: 'Space Grotesk', sans-serif; height: 40px; font-size: 16px; padding: 0 20px; border: none; cursor: pointer" onclick="window.open('https://drive.google.com/file/d/1QiJFOmMjvDgsDo3i1XzlQnImxe3RU_Cy/view?usp=sharing')">Conference Handbook</button></html>

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 2026". Please be sure to check your spam email folder if you do not see an email confirmation right away.

To access this event page, you are required to register.
Please complete your registration to continue.

We recommend reading [**the registration information**](https://2026.aclweb.org/registration/) first.

Registration Required

ACL 2026

The 64th Annual Meeting of the Association for Computational Linguistics (ACL 2026) will take place in San Diego, California, from July 2 - 7, 2026.

Language models (LMs) hallucinate. We inquire: Can we detect and mitigate hallucinations before they happen? This work answers this research question in the positive, by showing that the internal representations of LMs provide rich signals that can be used for this purpose. We introduce FactCheckmate, which preemptively detects hallucinations by learning a classifier that predicts whether the LM will hallucinate, based on the model's hidden states produced over the inputs, before decoding begins. If a hallucination is detected, FactCheckmate then intervenes by adjusting the LM's hidden states such that the model will produce more factual outputs. FactCheckmate provides fresh insights that the inner workings of LMs can be revealed by their hidden states. Practically, both its detection and mitigation models are lightweight, adding little inference overhead; FactCheckmate proves a more efficient approach for mitigating hallucinations compared to many post-hoc alternatives. We evaluate FactCheckmate over LMs of different scales and model families (including Llama, Mistral, Qwen and Gemma),across a variety of QA datasets from different domains. Our results demonstrate the effectiveness of FactCheckmate, achieving over 70% preemptive detection accuracy. On average, outputs generated by LMs with intervention are 34.4% more factual compared to those without.

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.

The biases exhibited by text-to-image (TTI) models are often treated as independent, though in reality, they may be deeply interrelated. Addressing bias along one dimension—such as ethnicity or age—can inadvertently affect another, like gender, either mitigating or exacerbating existing disparities. Understanding these interdependencies is crucial for designing fairer generative models, yet measuring such effects quantitatively remains a challenge. To address this, we introduce BiasConnect, a novel tool for analyzing and quantifying bias interactions in TTI models. BiasConnect uses counterfactual interventions along different bias axes to reveal the underlying structure of these interactions and estimates the effect of mitigating one bias axis on another. These estimates show strong correlation (+0.65) with observed post-mitigation outcomes. Building on BiasConnect, we propose InterMit, an intersectional bias mitigation algorithm guided by user-defined target distributions and priority weights. InterMit achieves lower bias (0.33 vs. 0.52) with fewer mitigation steps (2.38 vs. 3.15 average steps), and yields superior image quality compared to traditional techniques. Although our implementation is training-free, InterMit is modular and can be integrated with many existing debiasing approaches for TTI models, making it a flexible and extensible solution.

Mitigate One, Skew Another? Tackling Intersectional Biases in Text-to-Image Models

KV Cache is commonly used to accelerate LLM inference with long contexts, yet its high memory demand drives the need for cache compression. Existing compression methods, however, are largely heuristic and lack dynamic budget allocation. To address this limitation, we introduce a unified framework for cache compression by minimizing information loss in Transformer residual streams. Building on it, we analyze the layer attention output loss and derive a new metric to compare cache entries across heads, enabling layer-wise compression with dynamic head budgets. Additionally, by contrasting cross-layer information, we also achieve dynamic layer budgets. LAVa is the first unified strategy for cache eviction and dynamic budget allocation that, unlike prior methods, does not rely on training or the combination of multiple strategies. Experiments with four benchmarks (LongBench, Needle-In-A-Haystack, Ruler, and InfiniteBench) demonstrate its superiority over strong baselines. Moreover, our experiments reveal a new insight: dynamic layer budgets are crucial for generation tasks (e.g., code completion), while dynamic head budgets play a key role in extraction tasks (e.g., extractive QA). As a fully dynamic compression method, LAVa consistently maintains top performance across task types.

LAVa: Layer-wise KV Cache Eviction with Dynamic Budget Allocation

Recent speech-LLMs have shown impressive performance in tasks like transcription and translation, yet they remain limited in understanding the paralinguistic aspects of speech crucial for social and emotional intelligence. We propose CP-Bench, a benchmark for evaluating speech-LLMs on contextual paralinguistic reasoning the integration of verbal content with non-verbal cues like emotion and prosody. The benchmark includes two curated question answering (QA) datasets requiring both linguistic and empathetic understanding. We evaluate state-of-the-art speech-LLMs from both open and closed-source models and perform a comprehensive analysis across different question types. The top two models were further analyzed under temperature tuning to understand its effect on this task. Our benchmark reveals a key gap in existing evaluations and offers insights into building more context-aware and emotionally intelligent speech-capable LLMs.

Benchmarking Contextual and Paralinguistic Reasoning in Speech-LLMs: A Case Study with In-the-Wild Data

Pre-trained language models have achieved remarkable success across diverse applications but remain susceptible to spurious, concept-driven correlations that impair robustness and fairness. In this work, we introduce CURE, a novel and lightweight framework that systematically disentangles and suppresses conceptual shortcuts while preserving essential content information. Our method first extracts concept-irrelevant representations via a dedicated content extractor reinforced by a reversal network, ensuring minimal loss of task-relevant information. A subsequent controllable debiasing module employs contrastive learning to finely adjust the influence of residual conceptual cues, enabling the model to either diminish harmful biases or harness beneficial correlations as appropriate for the target task. Evaluated on the IMDB and Yelp datasets using three pre-trained architectures, CURE achieves an absolute improvement of +10 points in F1 score on IMDB and +2 points on Yelp, while introducing minimal computational overhead. Our approach establishes a flexible, unsupervised blueprint for combating conceptual biases, paving the way for more reliable and fair language understanding systems.

CURE: Controlled Unlearning for Robust Embeddings — Mitigating Conceptual Shortcuts in Pre-Trained Language Models

Figurative language and figures of speech, such as metaphors and hyperboles, are used every day in written and oral communication among human beings. Nonetheless, this imaginative use of words in a non literal way requires a solid understanding of semantics and a deep real-world knowledge. In the longstanding debate about whether Neural Language Models (NLMs) really have a full understanding of text, analysing how they can recognise figurative language can provide some intuition of their functioning, their capabilities and their limits. Therefore, in this paper, we exploit probing tasks to study how several NLMs of different sizes recognise four different figures of speech: hyperboles, metaphors, oxymorons and pleonasms. We analyse whether this information is learned and how it is acquired during the training of the model, describing its learning trajectory. Moreover, we analyse which layers have a better comprehension of figurative language and the influence of pre-training data.

Learning Trajectories of Figurative Language for Pre-Trained Language Models

Large language models (LLMs) have delivered significant breakthroughs across diverse domains but can still produce unreliable or misleading outputs, posing critical challenges for real-world applications. While many recent studies focus on quantifying model uncertainty, relatively little work has been devoted to \textit{diagnosing the source of uncertainty}. In this study, we show that, when an LLM is uncertain, the patterns of disagreement among its multiple generated responses contain rich clues about the underlying cause of uncertainty. To illustrate this point, we collect multiple responses from a target LLM and employ an auxiliary LLM to analyze their patterns of disagreement. The auxiliary model is tasked to reason about the likely source of uncertainty, such as whether it stems from ambiguity in the input question, a lack of relevant knowledge, or both. In cases involving knowledge gaps, the auxiliary model also identifies the specific missing facts or concepts contributing to the uncertainty. In our experiment, we validate our framework on AmbigQA, OpenBookQA, and MMLU-Pro, confirming its generality in diagnosing distinct uncertainty sources. Such diagnosis shows the potential for relevant manual interventions that improve LLM performance and reliability.

Can Multiple Responses from an LLM Reveal the Sources of Its Uncertainty?

In our era of widespread false information, human fact-checkers often face the challenge of duplicating efforts when verifying claims that may have already been addressed in other countries or languages. As false information transcends linguistic boundaries, the ability to automatically detect previously fact-checked claims across languages has become an increasingly important task. This paper presents the first comprehensive evaluation of large language models (LLMs) for multilingual previously fact-checked claim detection. We assess seven LLMs across 20 languages in both monolingual and cross-lingual settings. Our results show that while LLMs perform well for high-resource languages, they struggle with low-resource languages. Moreover, translating original texts into English proved to be beneficial for low-resource languages. These findings highlight the potential of LLMs for multilingual previously fact-checked claim detection and provide a foundation for further research on this promising application of LLMs.

Large Language Models for Multilingual Previously Fact-Checked Claim Detection

Code-switching presents a complex challenge for syntactic analysis, especially in low-resource language settings where annotated data is scarce. While recent work has explored the use of large language models (LLMs) for sequence-level tagging, few approaches systematically investigate how well these models capture syntactic structure in code-switched contexts. Moreover, existing parsers trained on monolingual treebanks often fail to generalize to multilingual and mixed-language input. To address this gap, we introduce the BiLingua Pipeline, an LLM-based annotation pipeline designed to produce Universal Dependencies (UD) annotations for code-switched text. First, we develop a prompt-based framework for Spanish-English and Spanish-Guaraní data, combining few-shot LLM prompting with expert review. Second, we release two annotated datasets, including the first Spanish-Guaraní UD-parsed corpus. Third, we conduct a detailed syntactic analysis of switch points across language pairs and communicative contexts. Experimental results show that BiLingua Pipeline achieves up to 95.29% LAS after expert revision, significantly outperforming prior baselines and multilingual parsers. These results show that LLMs, when carefully guided, can serve as practical tools for bootstrapping syntactic resources in under-resourced, code-switched environments.

Parsing the Switch: LLM-Based UD Annotation for Complex Code-Switched and Low-Resource Languages

Food safety demands timely detection, regulation, and public communication, yet the lack of structured datasets hinders NLP research. We present a new dataset of food safety documents with both human-written and LLM-generated summaries, plus metadata. We evaluate its utility on multilabel classification, document retrieval, and question answering via retrieval-augmented generation, showing that LLM summaries perform comparably or better than human ones. Clustering of summaries reveals potential for event tracking and compliance monitoring. This work supports AI-driven food safety applications in risk detection, policy enforcement, and public health.

FoodSafeSum: Enabling Natural Language Processing Applications for Food Safety Document Summarization and Analysis

As organizations increasingly rely on AI systems for decision support in sustainability contexts, it becomes critical to understand the inherent biases and perspectives embedded in Large Language Models (LLMs). This study systematically investigates how five state-of-the-art LLMs -- Claude, DeepSeek, GPT, LLaMA, and Mistral -- conceptualize sustainability and its relationship with AI. We administered validated, psychometric sustainability-related questionnaires -- each 100 times per model -- to capture response patterns and variability. Our findings revealed significant inter-model differences: For example, GPT exhibited skepticism about the compatibility of AI and sustainability, whereas LLaMA demonstrated extreme techno-optimism with perfect scores for several Sustainable Development Goals (SDGs). Models also diverged in attributing institutional responsibility for AI and sustainability integration, a results that holds implications for technology governance approaches. Our results demonstrate that model selection could substantially influence organizational sustainability strategies, highlighting the need for awareness of model-specific biases when deploying LLMs for sustainability-related decision-making.

Downloads

Next from EMNLP 2025

Mitigate One, Skew Another? Tackling Intersectional Biases in Text-to-Image Models

Similar lecture

Language Statistics and False Belief Reasoning: Evidence from 41 Open-Weight LMs