Austria

This work evaluates the potential of large language models (LLMs) to power digital assistants capable of complex action execution. Such assistants rely on pre-trained programming knowledge to execute multi-step goals by composing objects and functions defined in assistant libraries into action execution programs. To achieve this, we develop ASPERA, a framework comprising an assistant library simulation and a human-assisted LLM data generation engine. Our engine allows developers to guide LLM generation of high-quality tasks consisting of complex user queries, simulation state and corresponding validation programs, tackling data availability and evaluation robustness challenges. Alongside the framework we release Asper-Bench, an evaluation dataset of $250$ challenging tasks generated using ASPERA, which we use to show that program generation grounded in custom assistant libraries is a significant challenge to LLMs compared to dependency-free code generation.

ACL 2025

ASPERA: A Simulated Environment to Evaluate Planning for Complex Action Execution

nlp datasets

language resources

corpus creation

benchmarking

evaluation

### 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 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)

Large Language Models (LLMs) have demonstrated impressive performance in biomedical relation extraction, even in zero-shot scenarios. However, evaluating LLMs in this task remains challenging due to their ability to generate human-like text, often producing synonyms or abbreviations of gold-standard answers, making traditional automatic evaluation metrics unreliable. On the other hand, while human evaluation is more reliable, it is costly and time-consuming, making it impractical for real-world applications. This paper investigates the use of LLMs-as-the-Judge as an alternative evaluation method for biomedical relation extraction. We benchmark 8 LLMs as judges to evaluate the responses generated by 5 other LLMs across 3 biomedical relation extraction datasets. Unlike other text-generation tasks, we observe that LLM-based judges perform quite poorly (usually below 50% accuracy) in the biomedical relation extraction task. Our findings reveal that it happens mainly because relations extracted by LLMs do not adhere to any standard format. To address this, we propose structured output formatting for LLM-generated responses that helps LLM-Judges to improve their performance by about 15% (on average). We also introduce a domain adaptation technique to further enhance LLM-Judge performance by effectively transferring knowledge between datasets. We release both our human-annotated and LLM-annotated judgment data (36k samples in total) for public use here: https://github.com/tahmedge/llm_judge_biomedical_re.

Improving Automatic Evaluation of Large Language Models (LLMs) in Biomedical Relation Extraction via LLMs-as-the-Judge

This paper focuses on a new task of answering geographic reasoning questions based on the given image (called GeoVQA). Unlike traditional VQA tasks, GeoVQA asks for details about the image-related culture, landscape, etc. This requires not only the identification of the objects in the image, their properties and relations, but also the understanding of the geographic knowledge of the objects, such as location, transportation, landmark, cuisine, etc. This background knowledge does not explicitly appear in the image, nor is there an extra-textual description. Without this missing but necessary knowledge, it is difficult for existing matching-based methods to infer the correct answer. To tackle these challenges, we propose a new geographic reasoning framework for our task. We first analyze the image and describe its fine-grained content by text and keywords using a multi-modal retrieval augmented technique, so as to deduce an answer in a unified textual modality. Next, we retrieve the crucial geographic commonsense knowledge. To reduce the retrieval complexity, we design a dynamic method that can adaptively collect the relevant clues for each reasoning step. The step in the incorrect direction will be pruned according to some judgment criteria. The remaining steps can help us form a reasoning chain to derive a correct answer. Moreover, we create a large-scale dataset GVQA with 41,329 samples to conduct the evaluation. The results demonstrate the effectiveness of our approach.

Answering Complex Geographic Questions by Adaptive Reasoning with Visual Context and External Commonsense Knowledge

Multi-token prediction (MTP) is a recently proposed pre-training objective for language models. Rather than predicting only the next token (NTP), MTP predicts the next *k* tokens at each prediction step, using multiple prediction heads. MTP has shown promise in improving downstream performance, inference speed, and training efficiency, particularly for large models. However, prior work has shown that smaller language models (SLMs) struggle with the MTP objective. To address this, we propose a curriculum learning strategy for MTP training, exploring two variants: a forward curriculum, which gradually increases the complexity of the pre-training objective from NTP to MTP, and a reverse curriculum, which does the opposite. Our experiments show that the forward curriculum enables SLMs to better leverage the MTP objective during pre-training, improving downstream NTP performance and generative output quality, while retaining the benefits of self-speculative decoding. The reverse curriculum achieves stronger NTP performance and output quality, but fails to provide any self-speculative decoding benefits.

Pre-Training Curriculum for Multi-Token Prediction in Language Models

Large language models excel at problem-solving but often struggle with complex reasoning and factual accuracy. While chain-of-thought and retrieval-augmented generation help break down problems and retrieve knowledge, they still falter on challenging tasks like competitive programming due to frequent reasoning errors and irrelevant retrieval. To address this, we introduce Critic-guided planning with Retrieval-augmentation, CR-Planner, a novel framework that leverages fine-tuned critic models to guide both reasoning and retrieval processes through planning. CR-Planner iteratively selects and executes sub-goals, guided by critic models. A sub-goal critic identifies promising sub-goals from reasoning, query generation, and retrieval, while an execution critic evaluates outputs of sub-goal executions. We employ Monte Carlo Tree Search to collect data for critic training, allowing systematic exploration of action sequences and effective navigation toward the final answer. We evaluate CR-Planner on challenging domain-knowledge-intensive and reasoning-heavy tasks, including competitive programming, theorem-driven math reasoning, and complex domain retrieval problems. It significantly outperforms baselines, demonstrating effectiveness in both reasoning and retrieval.

Can We Further Elicit Reasoning in LLMs? Critic-Guided Planning with Retrieval-Augmentation for Solving Challenging Tasks

Legal scholars are increasingly using corpus based methods for assessing historical meaning. Among work focused on the so-called founding era (mid to late 18th century), the majority of such studies use the Corpus of Founding Era American English (COFEA) and rely on methods such as word counting and manual coding. Here, we demonstrate what can be inferred about meaning change and variation using more advanced NLP methods, focusing on terms in the U.S. Constitution. We also carry out a data quality assessment of COFEA, pointing out issues with OCR quality and metadata, compare diachronic change to synchronic variation, and discuss limitations when using NLP methods for studying historical meaning.

Meaning Variation and Data Quality in the Corpus of Founding Era American English

It has been frequently observed that human speakers align their language use with each other during conversations. In this paper, we study empirically whether large language models (LLMs) exhibit the same behavior of conversational adaptation.
We construct a corpus of conversations between LLMs and find that two LLM agents end up making more similar syntactic choices as conversations go on, confirming that modern LLMs adapt their language use to their conversational partners in at least a rudimentary way.

LLMs syntactically adapt their language use to their conversational partner

Do language models (LMs) offer insights into human language learning? A common argument against this idea is that because their architecture and training paradigm are so vastly different from humans, LMs can learn arbitrary inputs as easily as natural languages. 
We test this claim by training LMs to model impossible and typologically unattested languages.
Unlike previous work, which has focused exclusively on English, we conduct experiments on 12 languages from 4 language families with two newly constructed parallel corpora. Our results show that while GPT-2 small can largely distinguish attested languages from their impossible counterparts, it does not achieve perfect separation between all the attested languages and all the impossible ones. We further test whether GPT-2 small distinguishes typologically attested from unattested languages with different NP orders by manipulating word order based on Greenberg's Universal 20. We find that the model's perplexity scores do not distinguish attested vs. unattested word orders, while its performance on the generalization test does. These findings suggest that LMs exhibit some human-like inductive biases, though these biases are weaker than those found in human learners.

Anything Goes? A Crosslinguistic Study of (Im)possible Language Learning in LMs

In this paper, we propose a new data synthesis method called \textbf{LogicPro}, which leverages LeetCode-style algorithm \underline{Pro}blems and their corresponding \underline{Pro}gram solutions to synthesize Complex \underline{Logic}al Reasoning data in text format. First, we synthesize complex reasoning problems through source algorithm problems and test cases. Then, standard answers and intermediate variable outputs are obtained for each problem based on standard python solutions and test cases. Finally, with the guidance of code intermediate variables, we synthesize the text reasoning process for each reasoning problems. Through this method, we can synthesize data that is difficult, scalable, effective, and comes with golden standard answers and high-quality reasoning processes. As a result, with our 540K synthesized dataset constructed solely from 2,360 algorithm problems, our approach achieves significant improvements in multiple models for the datasets \textit{BBH^27}, \textit{LogicBench}, \textit{DROP}, \textit{AR-LSAT}, and \textit{GSM8K}, etc. outperforming a wide range of existing reasoning datasets.

LogicPro: Improving Complex Logical Reasoning via Program-Guided Learning

Improving cultural competence of language technologies is important. However most recent works rarely engage with the communities they study, and instead rely on synthetic setups and imperfect proxies of culture. In this work, we take a human-centered approach to discover and measure language-based cultural norms, and cultural competence of LLMs. We focus on a single kind of culture, *research cultures*, and a single task, *adapting writing across research cultures*. Through a set of interviews with interdisciplinary researchers, who are experts at moving between cultures, we create a framework of structural, stylistic, rhetorical, and citational norms that vary across research cultures. We operationalise these features with a suite of computational metrics and use them for (a) surfacing latent cultural norms in human-written research papers at scale; and (b) highlighting the lack of cultural competence of LLMs, and their tendency to homogenize writing. Overall, our work illustrates the efficacy of a human-centered approach to measuring cultural norms in human-written and LLM-generated texts.

Research Borderlands: Analysing Scientific Writing Across Research Cultures

Large language models (LLMs) have demonstrated remarkable capabilities in various domains, including radiology report generation. Previous approaches have attempted to utilize multimodal LLMs for this task, enhancing their performance through the integration of domain-specific knowledge retrieval. However, these approaches often overlook the knowledge already embedded within the LLMs, leading to redundant information integration. To address this limitation, we propose Radar, a framework for enhancing radiology report generation with supplementary knowledge injection. Radar improves report generation by systematically leveraging both the internal knowledge of an LLM and externally retrieved information. Specifically, it first extracts the model's acquired knowledge that aligns with expert image-based classification outputs. It then retrieves relevant supplementary knowledge to further enrich this information. Finally, by aggregating both sources, Radar generates more accurate and informative radiology reports. Extensive experiments on MIMIC-CXR, CheXpert-Plus, and IU X-ray demonstrate that our model outperforms state-of-the-art LLMs in both language quality and clinical accuracy

Downloads

Next from ACL 2025

Improving Automatic Evaluation of Large Language Models (LLMs) in Biomedical Relation Extraction via LLMs-as-the-Judge

Similar lecture

Debug like a Human: A Large Language Model Debugger via Verifying Runtime Execution Step by Step

Stay up to date with the latest Underline news!

PRODUCT

COMPANY

RESOURCES