Austria

Large language models (LLMs) have shown promise in formal theorem proving, but their token-level processing often fails to capture the inherent hierarchical nature of mathematical proofs. We introduce \textbf{Hierarchical Attention}, a regularization method that aligns LLMs&#39; attention mechanisms with mathematical reasoning structures. Our approach establishes a five-level hierarchy from foundational elements to high-level concepts, ensuring structured information flow in proof generation. Experiments demonstrate that our method improves proof success rates by 2.05% on miniF2F and 1.69% on ProofNet while reducing proof complexity by 23.81% and 16.50% respectively. The code and models will be available.

ACL 2025

Hierarchical Attention Generates Better Proofs

improving theorem proving performance while reducing proof complexity on both minif2f and proofnet benchmarks.

we propose hierarchical attention to align llms' attention with mathematical reasoning structures

### Welcome to 2025 Annual Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics

Welcome to the 2025 meeting of the Nations of the Americas Chapter of the Association for Computational Linguistics! I am proud to help organize the first NAACL conference to carry the new name of our organization, one that emphasizes inclusion for all of the Americas. I am also pleased to welcome you to Albuquerque, New Mexico, a state whose unique blend of cultural influences will make for an excellent backdrop for NAACL 2025, especially with this year’s special theme on NLP in a Multicultural World. 
**[Continue reading...](https://drive.google.com/file/d/1jX-qGhqVSZZCIrAnJaz798pflu5Irrdn/view?usp=sharing)**

*- Colin Cherry, Google, NAACL 2025 General Chair* 

[![](https://assets.underline.io/markdown_image/1/image/b087f8a4dc5816d6e1a6514e59c59ac3.png)](https://drive.google.com/file/d/1T96GzPqObXrMln2BMByCSXTSizjTg69P/view?usp=sharing)

You need to log in with the email address you registered with. Access credentials have been sent to your email. 

Please be sure to check your spam and other email folders if you do not see an email confirmation right away.

Please log in to explore this event.

To access NAACL 2025 event page you are required to register. Please follow [**this link**](https://2025.naacl.org/registration/registration/) to register. Access will depend on your registration type.

Please register!

NAACL 2025

Welcome to 2025 Annual Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics

Large language models (LLMs) have shown promise in formal theorem proving, but their token-level processing often fails to capture the inherent hierarchical nature of mathematical proofs. We introduce \textbf{Hierarchical Attention}, a regularization method that aligns LLMs' attention mechanisms with mathematical reasoning structures. Our approach establishes a five-level hierarchy from foundational elements to high-level concepts, ensuring structured information flow in proof generation. Experiments demonstrate that our method improves proof success rates by 2.05% on miniF2F and 1.69% on ProofNet while reducing proof complexity by 23.81% and 16.50% respectively. The code and models will be available.

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)

Self-critique mechanisms significantly improve the performance of language models in complex reasoning tasks by giving them the ability to correct errors, conduct induction and deduction, and switch thinking insights. However, synthetic data methods often require human-introduced errors or sampling of the model's reasoning results from the previous moment, and the current output distribution of the model cannot be obtained, makes the data for critique and reasoning face the problem of distribution shifts. In this work, we propose an on-policy reinforcement learning framework to synchronize the reasoning and critique capabilities of language models. To alleviate reward hacking caused by outcome-based supervision, we design a deliberate reward framework for different purposes. The reward framework not only supervises the model reasoning process based on the results, but also uses Monte Carlo sampling to give appropriate rewards to the critique content according to the success rate of the model's correction after critique. In addition, we introduce a rule-based reward function to impose penalties on the model when it generates hallucinatory critiques. When our approach is applied to the DeepSeek-Math-7B-Base and Qwen2.5-7B-Base models, model performance improves 5.40 and 3.66 points, respectively, compared to the best baseline approach. This validates the significant advantages of our method in improving model's reasoning and self-critique capability. Code will be made available at https://github.com/rbao2018/SCOP

Fixing Distribution Shifts of LLM Self-Critique via On-Policy Self-Play Training

Text editing frames grammatical error correction (GEC) as a sequence tagging problem, where edit tags are assigned to input tokens, and applying these edits results in the corrected text. This approach has gained attention for its efficiency and interpretability. However, while extensively explored for English, text editing remains largely underexplored for morphologically rich languages like Arabic. In this paper, we introduce a text editing approach that derives edit tags directly from data, eliminating the need for language-specific edits. We demonstrate its effectiveness on Arabic, a diglossic and morphologically rich language, and investigate the impact of different edit representations on model performance. Our approach achieves SOTA results on two Arabic GEC benchmarks and performs on par with SOTA on two others. Additionally, our models are over six times faster than existing Arabic GEC systems, making our approach more practical for real-world applications. Finally, we explore ensemble models, demonstrating how combining different models leads to further performance improvements. We make our code, data, and pretrained models publicly available.

Enhancing Text Editing for Grammatical Error Correction: Arabic as a Case Study

Generative large language models are crucial in natural language processing, but they are vulnerable to backdoor attacks, where subtle triggers compromise their behavior. Although backdoor attacks against LLMs are constantly emerging, existing benchmarks remain limited in terms of sufficient coverage of attack, metric system integrity, backdoor attack alignment. And existing pre-trained backdoor attacks are idealized in practice due to resource access constraints. Therefore we establish $\textit{ELBA-Bench}$, a comprehensive and unified framework that allows attackers to inject backdoor through parameter efficient fine-tuning ($\textit{e.g.,}$ LoRA) or without fine-tuning techniques ($\textit{e.g.,}$ In-context-learning). $\textit{ELBA-Bench}$ provides over 1300 experiments encompassing the implementations of 12 attack methods, 18 datasets, and 12 LLMs. Extensive experiments provide new invaluable findings into the strengths and limitations of various attack strategies. For instance, PEFT attack consistently outperform without fine-tuning approaches in classification tasks while showing strong cross-dataset generalization with optimized triggers boosting robustness; Task-relevant backdoor optimization techniques or attack prompts along with clean and adversarial demonstrations can enhance backdoor attack success while preserving model performance on clean samples. Additionally, we introduce a universal toolbox designed for standardized backdoor attack research at https://github.com/NWPUliuxx/ELBA_Bench, with the goal of propelling further progress in this vital area.

ELBA-Bench: An Efficient Learning Backdoor Attacks Benchmark for Large Language Models

Continually expanding new languages for existing large language models (LLMs) is a promising yet challenging approach to building powerful multilingual LLMs.
The biggest challenge is to make the model continuously learn new languages while preserving the proficient ability of old languages.
To achieve this, recent work utilizes the Mixture-of-Experts (MoE) architecture to expand new languages by adding new experts and avoid catastrophic forgetting of old languages by routing corresponding tokens to the original model backbone (old experts).
Although intuitive, this kind of method is parameter-costly when expanding new languages and still inevitably impacts the performance of old languages.
To address these limitations, we analyze the language characteristics of different layers in LLMs and propose a layer-wise expert allocation algorithm (LayerMoE) to determine the appropriate number of new experts for each layer.
Specifically, we find different layers in LLMs exhibit different representation similarities between languages and then utilize the similarity as the indicator to allocate experts for each layer, i.e., the higher similarity, the fewer experts.
Additionally, to further mitigate the forgetting of old languages, we add a classifier in front of the router network on the layers with higher similarity to guide the routing of old language tokens.
Experimental results show that our method outperforms the previous state-of-the-art baseline with 60\% fewer experts in the single-expansion setting and with 33.3\% fewer experts in the lifelong-expansion setting, demonstrating the effectiveness of our method.

Less, but Better: Efficient Multilingual Expansion for LLMs via Layer-wise Mixture-of-Experts

Large language models (LLMs) have demonstrated significant advancements in code generation, yet they still face challenges when tackling tasks that extend beyond their basic capabilities. Recently, the concept of self-debugging has been proposed as a way to enhance code generation performance by leveraging execution feedback from tests. However, the availability of high-quality tests in real-world scenarios is often limited. In this context, self-debugging with self-generated tests emerges as a promising solution, though its limitations and practical potential have not been fully explored. To address this gap, we investigate the efficacy of self-debugging in code generation tasks. We propose and analyze two distinct paradigms for the self-debugging process: post-execution and in-execution self-debugging. Our findings reveal that post-execution self-debugging struggles with the test bias introduced by self-generated tests, which can lead to misleading feedback. In contrast, in-execution self-debugging enables LLMs to mitigate this bias and leverage intermediate states during program execution. By focusing on runtime information rather than relying solely on potentially flawed self-generated tests, this approach demonstrates significant promise for improving the robustness and accuracy of LLMs in code generation tasks.

Revisit Self-Debugging with Self-Generated Tests for Code Generation

Automated Essay Scoring (AES) plays a crucial role in language assessment. In particular, cross-prompt essay trait scoring provides learners with valuable feedback to improve their writing skills. However, due to the scarcity of prompts, most existing methods overlook critical information, such as content from prompts or essays, resulting in incomplete assessment perspectives. In this paper, we propose a robust AES framework, the Mixture of Ordered Scoring Experts (MOOSE), which integrates information from both prompts and essays. MOOSE employs three specialized experts to evaluate (1) the overall quality of an essay, (2) the relative quality across multiple essays, and (3) the relevance between an essay and its prompt. MOOSE introduces the ordered aggregation of assessment results from these experts along with effective feature learning techniques. Experimental results demonstrate that MOOSE achieves exceptionally stable and state-of-the-art performance in both cross-prompt scoring and multi-trait scoring on the ASAP++ dataset. The source code is released at https://github.com/antslabtw/MOOSE-AES.

Mixture of Ordered Scoring Experts for Cross-prompt Essay Trait Scoring

Large language models face intrinsic limitations in coding with APIs that are unseen in their training corpora. As libraries continuously evolve, it becomes impractical to exhaustively retrain LLMs with new API knowledge. This limitation hampers LLMs from solving programming problems which require newly introduced or privately maintained libraries. Inspired by exploratory programming paradigm in human behavior, we propose **ExploraCoder**, a training-free framework that empowers LLMs to invoke multiple unseen APIs in code solution by (1) planning a complex problem into several API invocation subtasks, and (2) experimenting with correct API usage at intermediate steps through a novel chain-of-API-exploration. We conduct evaluation on program synthesizing tasks involving complex API interactions. Experimental results demonstrate that ExploraCoder significantly improves performance for models lacking prior API knowledge, achieving absolute increases of up to 11.99\% over retrieval-based approaches and 17.28\% over pretraining-based methods in pass@10.

ExploraCoder: Advancing Code Generation for Multiple Unseen APIs via Planning and Chained Exploration

Inspired by theories in language psychology, it is natural to consider more constraints, such as intentions, logic, knowledge, etc., when a complex or multi-hop question is generated. As the subtask of Multi-Hop Question Generation (MHQG), the task of Multi-Constraint Multi-Hop Question Generation (MCHQG) is more aligned with human question theories. However, it is hard to determine how to bring various high-dimensional semantic constraints, and how to integrate each constraint across all hops when a multi-hop question is being generating. To address these challenges, we introduce an effective framework which includes constraint dimensionality reduction and divide-and-conquer-based dynamic projection; we call it RUBY. The proposed RUBY contains a module of high-dimensional semantic constraint dimension reduction and a module of sub-question answer pairs-based multi-hop question generation. Meanwhile, a Reasoning Dynamic Projection strategy is tailored to effectively incorporate the constraints into every hop of the multi-hop question. The experimental results demonstrate that RUBY consistently outperforms baseline models, which suggest that RUBY is able to effectively capture and integrate semantic constraints, leading to more accurate and human-like multi-hop question generation. Our code and data are available.

RUBY: An Effective Framework for Multi-Constraint Multi-Hop Question Generation

Prompt injection attacks manipulate large language models (LLMs) by misleading them to deviate from the original input instructions and execute maliciously injected instructions, because of their instruction-following capabilities and inability to distinguish between the original input instructions and maliciously injected instructions. 
To defend against such attacks, recent studies have developed various detection mechanisms. 
If we restrict ourselves specifically to works which perform detection rather than direct defense, most of them focus on direct prompt injection attacks, while there are few works for the indirect scenario, where injected instructions are indirectly from external tools, such as a search engine. Moreover, current works mainly investigate injection detection methods and pay less attention to the post-processing method that aims to mitigate the injection after detection.
In this paper, we investigate the feasibility of detecting and removing indirect prompt injection attacks, and we construct a benchmark dataset for evaluation. For detection, we assess the performance of existing LLMs and open-source detection models, and we further train detection models using our crafted training datasets. For removal, we evaluate two intuitive methods: (1) the *segmentation removal method*, which segments the injected document and removes parts containing injected instructions, and (2) the *extraction removal method*, which trains an extraction model to identify and remove injected instructions.

Can Indirect Prompt Injection Attacks Be Detected and Removed?

This work introduces RARE (Retrieval-Augmented Reasoning Enhancement), a versatile extension to the mutual reasoning framework (rStar), aimed at enhancing reasoning accuracy and factual integrity across large language models (LLMs) for complex, knowledge-intensive tasks such as medical and commonsense reasoning. RARE incorporates two innovative actions within the Monte Carlo Tree Search (MCTS) framework: (A6), which generates search queries based on the initial problem statement, performs information retrieval using those queries, and augments reasoning with the retrieved data to formulate the final answer; and (A7), which leverages information retrieval specifically for generated sub-questions and re-answers these sub-questions with the relevant contextual information. Additionally, a Retrieval-Augmented Factuality Scorer is proposed to replace the original discriminator, prioritizing reasoning paths that meet high standards of factuality. Experimental results with LLaMA 3.1 show that RARE enables open-source LLMs to achieve competitive performance with top closed-source models like GPT-4 and GPT-4o. This research establishes RARE as a scalable solution for improving LLMs in domains where logical coherence and factual integrity are critical.

Downloads

Next from ACL 2025

Fixing Distribution Shifts of LLM Self-Critique via On-Policy Self-Play Training

Similar lecture

Automating Mathematical Proof Generation Using Large Language Model Agents and Knowledge Graphs

Stay up to date with the latest Underline news!

PRESENTATIONS

CONFERENCES

COMPANY

RESOURCES