Austria

This article examines LLMs’ ability to correctly label simple inferences with partisan conclusions. For this, we develop a dataset with both formal and material inferences, containing logically equivalent pairs of inferences with conclusions that favor either the political left or the political right. This allows us to focus on political bias as a source of decrease in performance. Our samples are synthetically generated and thus highly controlled, covering both English and German. We assess the performance of 16 configurations of both open and proprietary state-of-the-art LLMs on that dataset, finding generally unreliable performance as well as widespread political bias which, in the case of the English samples, persists throughout our experimental settings.

ACL 2025

Can You Trust LLMs&#39; Judgements of the Validity of Simple Inferences With Partisan Conclusions? -- No!

llms

political bias

natural language inference

argumentation

### 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.

Can You Trust LLMs' Judgements of the Validity of Simple Inferences With Partisan Conclusions? -- No!

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)

Lexical borrowing, the adoption of words from one language into another, is a ubiquitous linguistic phenomenon influenced by geopolitical, societal, and technological factors. This paper introduces ConLoan--a novel contrastive dataset comprising sentences with and without loanwords across 10 languages. Through systematic evaluation using this dataset, we investigate how state-of-the-art machine translation and language models process loanwords compared to their native alternatives. Our experiments reveal that these systems show systematic preferences for loanwords over native terms and exhibit varying performance across languages. These findings provide valuable insights for developing more linguistically robust NLP systems.

ConLoan: A Contrastive Multilingual Dataset for Evaluating Loanwords

Large language models (LLMs) have demonstrated strong effectiveness and robustness when fine-tuned as dense retrievers.
However, their large parameter size presents significant computational challenges at inference time.
While smaller retrievers offer better efficiency, they often fail to generalize effectively with limited supervised fine-tuning data.
In this work, we introduce DRAMA, a training framework that leverages LLMs to train smaller generalizable dense retrievers.
In particular, we adopt pruned LLMs as the backbone and train on diverse LLM-augmented data in a single-stage contrastive learning setup.
Experiments show that DRAMA offers better multilingual and long-context capabilities than traditional encoder-based retrievers, and achieves strong performance across multiple tasks and languages.

DRAMA: Diverse Augmentation from Large Language Models Towards Smaller Generalizable Dense Retrievers

Real-world open-domain questions can be complex, especially when answering them requires integrating information from multiple sources. Effectively identifying the necessary information involves *aligning* it with the available data and its organization. However, existing RAG solutions address the alignment problem in a limited manner. Using off-the-shelf LLMs for question decomposition lacks awareness of the available data and its structure, often resulting in suboptimal retrieval performance. Alternatively, iteratively generating follow-up queries and interacting with the data collection, as explored in agentic RAG approaches, shows potential but is often *inefficient* since each successive query depends on previous results rather than being guided by the overall organization of the available data. To address the *alignment* problem, we introduce an LLM-based retrieval method --- ARM, designed to better align questions with the organization of the data collection. Instead of solely matching query utterance, ARM explores *relationships among data objects*, enabling a retrieve-all-at-once solution for complex queries. Experimental results demonstrate that ARM significantly outperforms existing RAG methods on various complex open-domain QA tasks across multiple modalities, achieving superior retrieval performance and downstream accuracy while significantly lowering monetary costs.

Can we Retrieve Everything All at Once? ARM: An Alignment-Oriented LLM-based Retrieval Method

Existing studies on fairness are largely Western-focused, making them inadequate for culturally diverse countries such as India. To address this gap, we introduce INDIC-BIAS, a comprehensive India-centric benchmark designed to evaluate fairness of LLMs across 85 identity groups encompassing diverse castes, religions, regions, and tribes. We first consult domain experts to curate over 1,800 socio-cultural topics spanning behaviors and situations, where biases and stereotypes are likely to emerge. Grounded in these topics, we generate and manually validate 20,000 real-world scenario templates to probe LLMs for fairness. We structure these templates into three evaluation tasks: plausibility, judgment, and generation. Our evaluation of 14 popular LLMs on these tasks reveals strong negative biases against marginalized identities, with models frequently reinforcing common stereotypes. Additionally, we find that models struggle to mitigate bias even when explicitly asked to rationalize their decision. Our evaluation provides evidence of both allocative and representational harms that current LLMs could cause towards Indian identities, calling for a more cautious usage in practical applications. We release INDIC-BIAS as an open-source benchmark to advance research on benchmarking and mitigating biases and stereotypes in the Indian context.

FairI Tales: Evaluation of Fairness in Indian Contexts with a Focus on Bias and Stereotypes

We introduce Speech-based Intelligence Quotient (SIQ) as a new form of human cognition-inspired evaluation pipeline for voice understanding large language models (LLM_{Voice}), designed to assess their voice understanding ability. 
Moving beyond popular voice understanding metrics such as word error rate (WER), SIQ examines LLM_{Voice} across three cognitive levels motivated by Bloom’s Taxonomy: (1) Remembering (i.e., WER for verbatim accuracy); (2) Understanding (i.e., similarity of LLM's interpretations); and (3) Application (i.e., QA accuracy for simulating downstream tasks). We demonstrate that SIQ not only quantifies voice understanding abilities but also provides unified comparisons between cascaded methods (e.g., ASR-LLM) and end-to-end models, identifies annotation errors in existing benchmarks, and detects hallucinations in LLM_{Voice}. Our framework represents a first-of-its-kind intelligence examination that bridges cognitive principles with voice-oriented benchmarks, while exposing overlooked challenges in multi-modal training. Our code and data will be open source to encourage future studies.

SpeechIQ: Speech-Agentic Intelligence Quotient Across Cognitive Levels in Voice Understanding by Large Language Models

Procedural texts help AI enhance reasoning about context and action sequences. Transforming these into Semantic Role Labeling (SRL) improves understanding of individual steps by identifying predicate-argument structure like {verb,what,where/with}. Procedural instructions are highly elliptic, for instance, (i) add cucumber to the bowl and (ii) add sliced tomatoes, the second step’s where argument is inferred from the context, referring to where the cucumber was placed. Prior SRL benchmarks often miss implicit arguments, leading to incomplete understanding. To address this, we introduce Implicit-VidSRL, a dataset that necessitates inferring implicit and explicit arguments from contextual information in multimodal cooking procedures. Our proposed dataset benchmarks multimodal models’ contextual reasoning, requiring entity tracking through visual changes in recipes. We study recent multimodal LLMs and reveal that they struggle to predict implicit arguments of what and where/with from multi-modal procedural data given the verb. Lastly, we propose iSRL-Qwen2-VL, which achieves a 17% relative improvement in F1-score for what-implicit and a 14.7% for where/with-implicit semantic roles over GPT-4o.

Predicting Implicit Arguments in Procedural Video Instructions

Prompt injection attacks pose a critical threat to large language models (LLMs), enabling goal hijacking and data leakage. Prompt guard models, though effective in defense, suffer from over-defense—falsely flagging benign inputs as malicious due to trigger word bias. To address this issue, 
we introduce NotInject, an evaluation dataset that systematically measures over-defense across various prompt guard models. NotInject contains 339 benign samples enriched with trigger words common in prompt injection attacks, enabling fine-grained evaluation. Our results show that state-of-the-art models suffer from over-defense issues, with accuracy dropping close to random guessing levels (60%). To mitigate this, we propose PIGuard, a novel prompt guard model that incorporates a new training strategy, Mitigating Over-defense for Free (MOF), which significantly reduces the bias on trigger words. PIGuard demonstrates state-of-the-art performance on diverse benchmarks including NotInject, surpassing the existing best model by 30.4%, offering a robust and open-source solution for detecting prompt injection attacks. The code and datasets are released at https://github.com/leolee99/PIGuard.

InjecGuard: Benchmarking and Mitigating Over-defense in Prompt Injection Guardrail Models

3D visual grounding (3DVG) involves localizing entities in a 3D scene referred to by natural language text. Such models are useful for embodied AI and scene retrieval applications, which involve searching for objects or patterns using natural language descriptions. While recent works have focused on LLM-based scaling of 3DVG datasets, these datasets do not capture the full range of potential prompts which could be specified in the English language. To ensure that we are scaling up and testing against a useful and representative set of prompts, we propose a framework for linguistically analyzing 3DVG prompts and introduce Visual Grounding with Diverse Language in 3D (ViGiL3D), a diagnostic dataset for evaluating visual grounding methods against a diverse set of language patterns. We evaluate existing open-vocabulary 3DVG methods to demonstrate that these methods are not yet proficient in understanding and identifying the targets of more challenging, out-of-distribution prompts, toward real-world applications.

ViGiL3D: A Linguistically Diverse Dataset for 3D Visual Grounding

Large language models (LLMs) have demonstrated remarkable abilities in various natural language processing areas, but they demand high computation resources which limits their deployment in real-world. Distillation is one technique to solve this problem through either knowledge distillation or task distillation. Both distillation approaches train small models to imitate specific features of LLMs, but they all neglect basic reading education for small models on generic texts that are \emph{unrelated} to downstream tasks. In this paper, we propose basic reading distillation (BRD) which educates a small model to imitate LLMs basic reading behaviors, such as named entity recognition, question raising and answering, on each sentence. After such basic education, we apply the small model on various tasks including language inference benchmarks and BIG-bench tasks. It shows that the small model can outperform or perform comparable to over 20x bigger LLMs. Analysis reveals that BRD effectively influences the probability distribution of the small model, and has orthogonality to either knowledge distillation or task distillation.

Basic Reading Distillation

Large Vision-Language Models (LVLMs) have demonstrated remarkable performance across diverse tasks. Despite great success, recent studies show that LVLMs encounter substantial limitations when engaging with visual graphs. To study the reason behind these limitations, we propose VGCure, a comprehensive benchmark covering 22 tasks for examining the fundamental graph understanding and reasoning capacities of LVLMs. Extensive evaluations conducted on 14 LVLMs reveal that LVLMs are weak in basic graph understanding and reasoning tasks, particularly those concerning relational or structurally complex information. Based on this observation, we propose a structure-aware fine-tuning framework to enhance LVLMs with structure learning abilities through three self-supervised learning tasks. Experiments validate the effectiveness of our method in improving LVLMs' performance on fundamental and downstream graph learning tasks, as well as enhancing their robustness against complex visual graphs.

Downloads

Next from ACL 2025

ConLoan: A Contrastive Multilingual Dataset for Evaluating Loanwords

Similar lecture

Measuring Political Bias in Large Language Models: What Is Said and How It Is Said