China

When leveraging large language models (LLMs) for multiple-choice question answering (MCQA), no convention exists regarding how the space following the last colon should be tokenized. We highlight this as an example for unclean LLM evaluation that complicates both NLP research and model use in an interdisciplinary context: we observe accuracy differences of up to 6% due to this (seemingly irrelevant) tokenization variation as well as reshuffled model rankings – a practical concern for leaderboards reliant on this prompt style. Surprisingly, we are able to recommend one specific strategy – tokenizing the space *together* with the answer letter – as we observe consistent and statistically significant performance improvements. Additionally, it improves model calibration. Our findings underscore the importance of careful evaluation design and highlight the need for standardized, transparent evaluation protocols to ensure reliable and comparable results.

EMNLP 2025

Mind the Gap: A Closer Look at Tokenization for Multiple-Choice Question Answering with LLMs

multiple-choice question answering

tokenization

large language models

evaluation

### 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!
"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.

Prompt sensitivity, referring to the phenomenon where minor variations in phrasing lead to significant changes in large language model (LLM) performance, has been widely accepted as a core limitation of LLMs. In this work, we revisit this issue and ask: Is the widely reported high prompt sensitivity truly an inherent weakness of LLMs, or is it largely an artifact of evaluation processes? To answer this question, we systematically evaluate 7 LLMs (e.g., GPT and Gemini family) across 6 benchmarks, including both multiple-choice and open-ended tasks on 12 diverse prompt templates. We find that much of the prompt sensitivity stems from heuristic evaluation methods, including log-likelihood scoring and rigid answer matching, which often overlook semantically correct responses expressed through alternative phrasings, such as synonyms or paraphrases. When we adopt LLM-as-judge evaluations, we observe a substantial reduction in performance variance and a consistently higher correlation in model rankings across prompts. Our findings suggest that modern LLMs are more robust to prompt templates than previously believed, and that prompt sensitivity may be more an artifact of evaluation than a flaw in the models.

Flaw or Artifact? Rethinking Prompt Sensitivity in Evaluating LLMs

Large language models (LLMs) can exhibit biases in reasoning capabilities due to linguistic modality, performing better on tasks in one language versus another, even with similar content. Most previous works evaluate this through reasoning tasks where reliance on strategies or knowledge can ensure success, such as in commonsense or math tasks. However, abstract reasoning is vital to reasoning for everyday life, where people apply "out-of-the-box thinking" to identify and use patterns for solutions, without a reliance on formulaic approaches. Comparatively, little work has evaluated linguistic biases in this task type. In this paper, we propose a task inspired by the New York Times Connections: GlobalGroup, that evaluates models in an abstract reasoning task across several languages. We constructed a game benchmark with five linguistic backgrounds – English, Spanish, Chinese, Hindi, and Arabic – in both the native language and an English translation for comparison. We also proposed game difficulty measurements to evaluate models on games with similar difficulty, enabling a more controlled comparison, which is particularly important in reasoning evaluations. Through experimentation, we find English modalities largely lead to better performance in this abstract reasoning task, and performance disparities between open- and closed-source models.

Think Globally, Group Locally: Evaluating LLMs Using Multi-Lingual Word Grouping Games

Mathematical reasoning remains a significant challenge for large language models (LLMs), despite progress in prompting techniques such as Chain-of-Thought (CoT). We present **Chain of Mathematically Annotated Thought (CoMAT)**, which enhances reasoning through two stages: *Symbolic Conversion* (converting natural language queries into symbolic form) and *Reasoning Execution* (deriving answers from symbolic representations). CoMAT operates entirely with a single LLM and without external solvers. Across four LLMs, CoMAT outperforms traditional CoT on six out of seven benchmarks, achieving gains of 4.48% on MMLU-Redux (MATH) and 4.58% on GaoKao MCQ. In addition to improved performance, CoMAT ensures faithfulness and verifiability, offering a transparent reasoning process for complex mathematical tasks.

CoMAT: Chain of Mathematically Annotated Thought Improves Mathematical Reasoning

Promises made by politicians, corporate leaders, and public figures have a significant impact on public perception, trust, and institutional reputation. However, the complexity and volume of such commitments, coupled with difficulties in verifying their fulfillment, necessitate innovative methods for assessing their credibility. This paper introduces the concept of Promise Verification, a systematic approach involving steps such as promise identification, evidence assessment, and the evaluation of timing for verification. We propose the first multilingual dataset, ML-Promise, which includes English, French, Chinese, Japanese, and Korean, aimed at facilitating in-depth verification of promises, particularly in the context of Environmental, Social, and Governance (ESG) reports. Given the growing emphasis on corporate environmental contributions, this dataset addresses the challenge of evaluating corporate promises, especially in light of practices like greenwashing. Our findings also explore textual and image-based baselines, with promising results from retrieval-augmented generation (RAG) approaches. This work aims to foster further discourse on the accountability of public commitments across multiple languages and domains.

ML-Promise: A Multilingual Dataset for Corporate Promise Verification

Zero-shot Event Detection (ED), the task of identifying event mentions in natural language text without any training data, is critical for document understanding in specialized domains. Understanding the complex event ontology, extracting domain-specific triggers from the passage, and structuring them appropriately overloads and limits the utility of Large Language Models (LLMs) for zero-shot ED. To this end, we propose DiCoRe, a divergent-convergent reasoning framework that decouples the task of ED using Dreamer and Grounder. Dreamer encourages divergent reasoning through open-ended event discovery, which helps to boost event coverage. Conversely, Grounder introduces convergent reasoning to align the free-form predictions with the task-specific instructions using finite-state machine guided constrained decoding. Additionally, an LLM-Judge verifies the final outputs to ensure high precision. Through extensive experiments on six datasets across five domains and nine LLMs, we demonstrate how DiCoRe consistently outperforms prior zero-shot, transfer-learning, and reasoning baselines, achieving 4–7% average F1 gains over the best baseline -- establishing DiCoRe as a strong zero-shot ED framework.

DiCoRe: Enhancing Zero-shot Event Detection via Divergent-Convergent LLM Reasoning

Large Language Models (LLMs) are widely used for temporal prediction, but their reliance on pretraining data raises contamination concerns, as accurate predictions on pre-cutoff test data may reflect memorization rather than reasoning, leading to an overestimation of their generalization capability. With the recent emergence of prompting-based unlearning techniques, a natural question arises: Can LLMs be prompted to simulate an earlier knowledge cutoff? In this work, we investigate the capability of prompting to simulate earlier knowledge cutoff in LLMs. We construct three evaluation datasets to assess the extent to which LLMs can forget (1) direct factual knowledge, (2) semantic shifts, and (3) causally related knowledge. Results demonstrate that while prompt-based simulated knowledge cutoffs show effectiveness when directly queried with the information after that date, they struggle to induce forgetting when the forgotten content is not directly asked but causally related to the query. These findings highlight the need for more rigorous evaluation settings when applying LLMs for temporal prediction tasks. Our dataset is included in the supplementary data file provided with this submission.

Can Prompts Rewind Time for LLMs? Evaluating the Effectiveness of Prompted Knowledge Cutoffs

Recent frontier-level LLMs have saturated many previously difficult benchmarks, leaving little room for further differentiation. This progress highlights the need for challenging benchmarks that provide objective verification. In this paper, we introduce MCBench, a benchmark designed to evaluate whether LLMs can execute string-matching NLP metrics by strictly following step-by-step instructions. Unlike prior benchmarks that depend on subjective judgments or general reasoning, MCBench offers an objective, deterministic and code-verifiable evaluation. This setup allows us to systematically test whether LLMs can maintain accurate step-by-step execution, including instruction adherence, numerical computation, and long-range consistency in handling intermediate results. To ensure objective evaluation of these abilities, we provide a parallel reference code that can evaluate the accuracy of LLM output. We provide three evaluative metrics and three benchmark variants designed to measure the detailed instruction understanding capability of LLMs. Our analyses show that MCBench serves as an effective and objective tool for evaluating the capabilities of cutting-edge LLMs

Metric Calculating Benchmark: Code-Verifiable Complicate Instruction Following Benchmark for Large Language Models

Large language models (LLMs) can now access a wide range of external tools, thanks to the Model Context Protocol (MCP). This greatly expands their abilities as various agents. However, LLMs rely entirely on the text descriptions of tools to decide which ones to use—a process that is surprisingly fragile. In this work, we expose a vulnerability in prevalent tool/function-calling protocols by investigating a series of edits to tool descriptions, some of which can drastically increase a tool's usage from LLMs when competing with alternatives. Through controlled experiments, we show that tools with properly edited descriptions receive **over 10 times more usage** from GPT-4.1 and Qwen2.5-7B than tools with original descriptions. We further evaluate how various edits to tool descriptions perform when competing directly with one another and how these trends generalize or differ across a broader set of 10 different models. These phenomenons, while giving developers a powerful way to promote their tools, underscore the need for a more reliable foundation for agentic LLMs to select and utilize tools and resources.

Tool Preferences in Agentic LLMs are Unreliable

Recent advances in large language models (LLMs) have yielded impressive gains on mathematical reasoning benchmarks via supervised fine-tuning (SFT). However, the brittleness of these models under input perturbations has cast doubt on whether such improvements reflect genuine reasoning abilities or merely superficial alignment with expected output formats. We investigate the mechanisms behind SFT improvements in small-scale LLMs, addressing four key questions: (1) Are performance gains primarily due to format alignment rather than reasoning? (2) Can high-quality supervision encourage genuine reasoning? (3) Does scaling data shift learning from format alignment to deeper reasoning? (4) Are format alignment gains consistent across model sizes and architectures? Through controlled experiments, we find that most performance improvements arise from format alignment rather than genuine reasoning enhancement. Moreover, SFT's effectiveness is strongly influenced by the alignment between the base model’s inductive biases and the teacher model’s output distribution, rather than the teacher’s raw strength. Finally, scaling up training data offers diminishing returns and does not fundamentally alter the model’s reasoning behavior. These findings suggest that current SFT practices may overestimate the reasoning abilities of LLMs and underscore the need for more rigorous evaluation methods.

The Emperor's New Reasoning: Format Imitation Overshadows Genuine Mathematical Understanding in SFT

Many statistical facts are conveyed through charts. While various methods have emerged for chart understanding, chart generation typically requires users to manually input code, intent, and other parameters to obtain the desired format on chart generation tools. Recently, the advent of image-generating Large Language Models has facilitated chart generation; however, even this process often requires users to provide numerous constraints for accurate results. In this paper, we propose a loop-based framework for automatically evolving charts in a multi-agent environment. Within this framework, three distinct agents—Chart Code Generator, Chart Replier, and Chart Quality Evaluator—collaborate for iterative, user-tailored chart generation using large language models. Our approach demonstrates an improvement of up to 29.97% in performance compared to first generation, while also reducing generation time by up to 86.9% compared to manual prompt-based methods, showcasing the effectiveness of this multi-agent collaboration in enhancing the quality and efficiency of chart generation.

Downloads

Next from EMNLP 2025

Flaw or Artifact? Rethinking Prompt Sensitivity in Evaluating LLMs

Similar lecture

Right Answer, Wrong Score: Uncovering the Inconsistencies of LLM Evaluation in Multiple-Choice Question Answering