Mexico

The growing awareness of safety concerns in large language models (LLMs) has sparked considerable interest in the evaluation of safety. This study investigates an under-explored issue about the evaluation of LLMs, namely the substantial discrepancy in performance between multiple-choice questions and open-ended questions. Inspired by research on jailbreak attack patterns, we argue this is caused by mismatched generalization. That is, LLM only remembers the answer style for open-ended safety questions, which makes it unable to solve other forms of safety tests. We refer to this phenomenon as fake alignment and construct a comparative benchmark to empirically verify its existence in LLMs. We introduce a Fake alIgNment Evaluation (FINE) framework and two novel metrics——Consistency Score (CS) and Consistent Safety Score (CSS), which jointly assess two complementary forms of evaluation to quantify fake alignment and obtain corrected performance estimation. Applying FINE to 14 widely-used LLMs reveals several models with purported safety are poorly aligned in practice. Subsequently, we found that multiple-choice format data can also be used as high-quality contrast distillation-based fine-tuning data, which can strongly improve the alignment consistency of LLMs with minimal fine-tuning overhead. For data and code, see https://github.com/AIFlames/Fake-Alignment.

NAACL 2024

Fake Alignment: Are LLMs Really Aligned Well?

ai alignment

ai safety

large language model

### 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**. 

ACL 2025

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

poster

## Welcome to 2024 Annual Conference of the North American Chapter of the Association for Computational Linguistics!
This year, the conference is in Mexico City. NAACL was actually already planned for Mexico City in 2021, but due to the pandemic the entire conference was moved online. This year, finally, we get to go! So it is my sincere pleasure to welcome you to Mexico City, whether in person or virtually. Having the conference in Mexico City is a good opportunity to emphasize that NAACL is our flagship conference for ACL members not only in North America but also in Central and South America, even though NAACL has been bearing “North” in its name. At this year’s conference, we have a theme to match, with a theme track on the Languages of Latin America to showcase the linguistic diversity of the region.
 
The opportunity to present at NAACL should not depend on a researcher’s travel budget, or their family status. This is why it is so important to make virtual participation at NAACL as good an experience as possible – but we want to also provide a good experience for in-person participants. As a community, we are still working out the best way to do that. This year at NAACL, we are trying out a big virtual poster session ahead of the conference, with the hope that this will make make for a lively and interactive experience. At the same time, we are reducing virtual oral presentations, which seem to be particularly tricky to make to work well. A big thanks to the NAACL program chairs and to Luciana Benotti for all their ideas and work to improve the virtual experience. And participants, virtual as well as in-person: Please let us know what worked for you and what didn’t, so we can continue to improve hybrid conferences. 
 
I have been lucky to work with many amazing people. Without their insight, dedication and patience, and without the many hours of work they put in, NAACL would not have been possible. A huge thank you to the program chairs Helena Gomez, Kevin Duh, and Steve Bethard – you are the best! 
 
Finally, I would like to thank all authors, invited speakers and panelists, area chairs and reviewers, the volunteers organizing and chairing sessions, and all attendees, in-person and virtual. Thank you for helping us make NAACL 2024 come to life. 
 
Welcome and hope you all enjoy the conference! 
Katrin Erk 
The University of Texas at Austin 
NAACL-2024 General Chair 
*You can read the full Welcome message in the [Conference Handbook (downloadable)](https://drive.google.com/file/d/1H1NvW0VASQjkSCYgw3mr-3l4yYDSxz6B/view?usp=sharing)*

To access the event page you need to register [**here.**](http://acl.swoogo.com/naacl2024) 
Your access to the event page is limited based on your registration type. If you registered for workshops only, you will gain access to full workshops content on the day of the workshops program.

Please register!

2024 Annual Conference of the North American Chapter of the Association for Computational Linguistics

Prior study shows that pre-training techniques can boost the performance of visual document understanding (VDU), which typically requires models to gain abilities to perceive and reason both document texts and layouts (e.g., locations of texts and table-cells). To this end, we propose visually guided generative text-layout pre-training, named ViTLP. Given a document image, the model optimizes hierarchical language and layout modeling objectives to generate the interleaved text and layout sequence. In addition, to address the limitation of processing long documents by Transformers, we introduce a straightforward yet effective multi-segment generative pre-training scheme, facilitating ViTLP to process word-intensive documents of any length. ViTLP can function as a native OCR model to localize and recognize texts of document images. Besides, ViTLP can be effectively applied to various downstream VDU tasks. Extensive experiments show that ViTLP achieves competitive performance over existing baselines on benchmark VDU tasks, including information extraction, document classification, and document question answering.

Visually Guided Generative Text-Layout Pre-training for Document Intelligence

Aspect-based sentiment analysis (ABSA) is a task that aims to determine the sentiment polarity of aspects by identifying opinion words. Recent advancements have predominantly been rooted either in semantic or syntactic methods. However, both of them tend to interference from local factors such as irrelevant words and edges, hindering the precise identification of opinion words. In this paper, we present Distance-based and Aspect-oriented Graph Convolutional Network (DAGCN) to address the aforementioned issue. Firstly, we introduce the Distance-based Syntactic Weight (DSW). It focuses on the local scope of aspects in the pruned dependency trees, thereby reducing the candidate pool of opinion words. Additionally, we propose Aspect-Fusion Attention (AF) to further filter opinion words within the local context and consider cases where opinion words are distant from the aspect. With the combination of DSW and AF, we achieve precise identification of corresponding opinion words. Extensive experiments on three public datasets demonstrate that the proposed model outperforms state-of-the-art models and verify the effectiveness of the proposed architecture.

DAGCN: Distance-based and Aspect-oriented Graph Convolutional Network for Aspect-based Sentiment Analysis

Existing self-supervised methods in natural language processing (NLP), especially hierarchical text classification (HTC), mainly focus on self-supervised contrastive learning, extremely relying on human-designed augmentation rules to generate contrastive samples, which can potentially corrupt or distort the original information. In this paper, we tend to investigate the feasibility of a contrastive learning scheme in which the semantic and syntactic information inherent in the input sample is adequately reserved in the contrastive samples and fused during the learning process. Specifically, we propose an information lossless contrastive learning strategy for HTC, namely $\textbf{H}$ierarchy-aware $\textbf{I}$nformation $\textbf{L}$ossless contrastive $\textbf{L}$earning (HILL), which consists of a text encoder representing the input document, and a structure encoder directly generating the positive sample. The structure encoder takes the document embedding as input, extracts the essential syntactic information inherent in the label hierarchy with the principle of structural entropy minimization, and injects the syntactic information into the text representation via hierarchical representation learning. Experiments on three common datasets are conducted to verify the superiority of HILL.

HILL: Hierarchy-aware Information Lossless Contrastive Learning for Hierarchical Text Classification

We study the patent phrase similarity inference task, which measures the semantic similarity between two patent phrases. As patent documents employ legal and highly technical language, existing semantic textual similarity methods that use localized contextual information do not perform satisfactorily in inferring patent phrase similarity. To address this, we introduce a graph-augmented approach to amplify the global contextual information of the patent phrases. For each patent phrase, we construct a phrase graph that links to its focal patents and a list of patents that are either cited by or cite these focal patents. The augmented phrase embedding is then derived from combining its localized contextual embedding with its global embedding within the phrase graph. We further propose a self-supervised learning objective that capitalizes on the retrieved topology to refine both the contextualized embedding and the graph parameters in an end-to-end manner. Experimental results from a unique patent phrase similarity dataset demonstrate that our approach significantly enhances the representation of patent phrases, resulting in marked improvements in similarity inference in a self-supervised fashion. Substantial improvements are also observed in the supervised setting, underscoring the potential benefits of leveraging retrieved phrase graph augmentation.

Connecting the Dots: Inferring Patent Phrase Similarity with Retrieved Phrase Graphs

Argument mining has focused so far mainly on the identification, extraction, and formalization of arguments. An important yet unaddressed task consists in the prediction of the argumentative behavior of stakeholders in a debate. Predicting the argumentative behavior in advance can support foreseeing issues in public policy making or help recognize potential disagreements early on and help to resolve them. In this paper, we consider the novel task of predicting the argumentative behavior of individual stakeholders. We present ARGENST, a framework that relies on a recommender-based architecture to predict the stance and the argumentative main point on a specific controversial topic for a given stakeholder, which is described in terms of a profile including properties related to demographic attributes, religious and political orientation, socio-economic background, etc. We evaluate our approach on the well-known debate.org dataset in terms of accuracy for predicting stance as well as in terms of similarity of the generated arguments to the ground truth arguments using BERTScore. As part of a case study, we show how juries of members representing different stakeholder groups and perspectives can be assembled to simulate the public opinion on a given topic.

“Tell me who you are and I tell you how you argue”: Predicting Stances and Arguments for Stakeholder Groups

Opinion summarization is automatically generating summaries from a variety of subjective information, such as product reviews or political opinions. The challenge of opinions summarization lies in presenting divergent or even conflicting opinions. We conduct an analysis of previous summarization models, which reveals their inclination to amplify the polarity bias, emphasizing the majority opinions while ignoring the minority opinions. To address this issue and make the summarizer express both sides of opinions, we introduce the concept of polarity calibration, which aims to align the polarity of output summary with that of input text. Specifically, we develop a reinforcement training approach for polarity calibration. This approach feeds the polarity distance between output summary and input text as reward into the summarizer, and also balance polarity calibration with content preservation and language naturality. We evaluate our Polarity Calibration model (PoCa) on two types of opinions summarization tasks: summarizing product reviews and political opinions articles. Automatic and human evaluation demonstrate that our approach can mitigate the polarity mismatch between output summary and input text, as well as maintain the content semantic and language quality.

Polarity Calibration for Opinion Summarization

Media outlets are becoming more partisan and polarized nowadays. In this paper, we identify media bias at the sentence level, and pinpoint bias sentences that intend to sway readers' opinions. As bias sentences are often expressed in a neutral and factual way, considering broader context outside a sentence can help reveal the bias. In particular, we observe that events in a bias sentence need to be understood in associations with other events in the document. Therefore, we propose to construct an event relation graph to explicitly reason about event-event relations for sentence-level bias identification. The designed event relation graph consists of events as nodes and four common types of event relations: coreference, temporal, causal, and subevent relations. Then, we incorporate event relation graph for bias sentences identification in two steps: an event-aware language model is built to inject the events and event relations knowledge into the basic language model via soft labels; further, a relation-aware graph attention network is designed to update sentence embedding with events and event relations information based on hard labels. Experiments on two benchmark datasets demonstrate that our approach with the aid of event relation graph improves both precision and recall of bias sentence identification.

Sentence-level Media Bias Analysis with Event Relation Graph

Grapheme-to-phoneme conversion (G2P) is a critical component of the text-to-speech system (TTS), where polyphone disambiguation is the most crucial task. However, polyphone disambiguation datasets often suffer from the long-tail problem, and context learning for polyphonic characters commonly stems from a single dimension. In this paper, we propose a novel model DLM: a Decoupled Learning Model for long-tailed polyphone disambiguation in Mandarin. Firstly, DLM decouples representation and classification learnings. It can apply different data samplers for each stage to obtain an optimal training data distribution. This can mitigate the long-tail problem. Secondly, two improved attention mechanisms and a gradual conversion strategy are integrated into the DLM, which achieve transition learning of context from local to global. Finally, to evaluate the effectiveness of DLM, we construct a balanced polyphone disambiguation corpus via in-context learning. Experiments on the benchmark CPP dataset demonstrate that DLM achieves a boosted accuracy of 99.07%. Moreover, DLM improves the disambiguation performance of long-tailed polyphonic characters. For many long-tailed characters, DLM even achieves an accuracy of 100%.

DLM: A Decoupled Learning Model for Long-tailed Polyphone Disambiguation in Mandarin

Prompt Engineering has garnered significant attention for enhancing the performance of large language models across a multitude of tasks. Techniques such as the Chain-of-Thought not only bolster task performance but also delineate a clear trajectory of reasoning steps, offering a tangible form of explanation for the audience. Prior works on interpretability assess the reasoning chains yielded by Chain-of-Thought solely along a singular axis, namely faithfulness. We present a comprehensive and multifaceted evaluation of interpretability, examining not only faithfulness but also robustness and utility across multiple commonsense reasoning benchmarks. Likewise, our investigation is not confined to a single prompting technique; it expansively covers a multitude of prevalent prompting techniques employed in large language models, thereby ensuring a wide-ranging and exhaustive evaluation. In addition, we introduce a simple interpretability alignment technique, termed Self-Entailment-Alignment Chain-of-thought, that yields more than 70\% improvements across multiple dimensions of interpretability. Code is available at https://github.com/SenticNet/CoT_interpretability

How Interpretable are Reasoning Explanations from Prompting Large Language Models?

Large-scale pre-trained language models have displayed unrivaled capacity in generating text that closely resembles human-written text. Nevertheless, generating texts adhering to specific conditions without fine-tuning or adding new parameters can be challenging. Contemporary approaches commonly rely on either prompts or auxiliary models to avoid modifying the language models. These auxiliary models are designed to assess whether a generated token contributes to meeting the desired requirements. These approaches adjust the distribution of the next token during the inference phase by leveraging the prediction score of the desired attribute to calculate gradients. However, these auxiliary models typically require the language model's latent states. This prerequisite challenges integrating various existing black box attribute models or tools. We present the Plug-in Language Model (PiLM) as a solution to address the limitations. PiLM leverages reinforcement learning to utilize black box tools directly, adjusting the latent state to control text generation. However, performing backpropagation during the inference phase is time-consuming for PiLM. By replacing backpropagation with a simple regression model, PiLM can achieve an inference time comparable to that of the original LLM. Experiment results show that our approaches in this paper outperform existing state-of-the-art methods that rely on gradient-based, weighted decoding, or prompt-based methodologies.

Downloads

Next from NAACL 2024

Visually Guided Generative Text-Layout Pre-training for Document Intelligence

Similar lecture

LSSF: Safety Alignment for Large Language Models through Low-Rank Safety Subspace Fusion

.css-70qvj9{display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;}Downloads

Next from NAACL 2024

Visually Guided Generative Text-Layout Pre-training for Document Intelligence

Similar lecture

LSSF: Safety Alignment for Large Language Models through Low-Rank Safety Subspace Fusion

Downloads