Austria

New LLM benchmarks are important to align with the rapid development of Large Language Models (LLMs). In this work, we present Chinese SimpleQA, the first comprehensive Chinese benchmark to evaluate the factuality ability of LLMs to answer short questions, and Chinese SimpleQA mainly has five properties (i.e., Chinese, Diverse, High-quality, Static, Easy-to-evaluate). Specifically, first, we focus on the Chinese language over 6 major topics with 99 diverse subtopics. Second, we conduct a comprehensive quality control process to achieve high-quality questions and answers, where the reference answers are static and cannot be changed over time. Third, following SimpleQA, the questions and answers are very short, and the grading process is easy-to-evaluate. Based on Chinese SimpleQA, we perform a comprehensive evaluation of the factuality abilities of existing LLMs. Finally, we hope that Chinese SimpleQA could guide the developers to better understand the Chinese factuality abilities of their models and facilitate the growth of LLMs.

ACL 2025

Chinese SimpleQA: A Chinese Factuality Evaluation for Large Language Models

large language models

factuality

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**. 

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

poster

Visual persuasion, which uses visual elements to influence cognition and behaviors, is crucial in fields such as advertising and political
communication. With recent advancements in artificial intelligence, there is growing potential to develop persuasive systems that automatically generate persuasive images tailored to individuals. However, a significant bottleneck in this area is the lack of comprehensive
datasets that connect the persuasiveness of images with the personal information about those who evaluated the images. To address this gap and facilitate technological advancements in personalized visual persuasion, we release the Personalized Visual Persuasion (PVP) dataset, comprising 28,454 persuasive images across 596 messages and 9 persuasion strategies. Importantly, the PVP dataset provides persuasiveness scores of images evaluated by 2,521 human annotators, along with their demographic and psychological characteristics (personality traits and values). We demonstrate the utility of our dataset by developing a persuasive image generator and an automated evaluator, and establish benchmark baselines. Our experiments reveal that incorporating psychological characteristics enhances the generation and evaluation of persuasive images, providing valuable insights for personalized visual persuasion.

PVP: An Image Dataset for Personalized Visual Persuasion with Persuasion Strategies, Viewer Characteristics, and Persuasiveness Ratings

With the popularity of multimodal techniques, it receives growing interests to acquire useful information in visual forms. In this work, we formally define an emerging IR paradigm called Visualized Information Retrieval, or Vis-IR, where multimodal information, such as texts, images, tables and charts, is jointly represented by a unified visual format called Screenshots, for various retrieval applications. We further make three key contributions for Vis-IR. First, we create VIRA (Vis-IR Aggregation), a large-scale dataset comprising a vast collection of screenshots from diverse sources, carefully curated into captioned and question-answer formats. Second, we develop UniSE (Universal Screenshot Embeddings), a family of retrieval models that enable screenshots to query or be queried across arbitrary data modalities. Finally, we construct MVRB (Massive Visualized IR Benchmark), a comprehensive benchmark covering a variety of task forms and application scenarios. Through extensive evaluations on MVRB, we highlight the deficiency from existing multimodal retrievers and the substantial improvements made by UniSE. Our data, model and benchmark have been made publicly available, which lays a solid foundation for this emerging field.

Any Information Is Just Worth One Single Screenshot: Unifying Search With Visualized Information Retrieval

Inference-time alignment methods have gained significant attention for their efficiency and effectiveness in aligning large language models (LLMs) with human preferences. However, existing dominant approaches, reward-guided search (RGS), suffer from a critical granularity mismatch: reward models (RMs) are trained on complete responses but applied to incomplete sequences during generation, leading to inconsistent scoring and suboptimal alignment. To combat the challenge, we argue that an ideal RM should satisfy two objectives: Score Consistency, ensuring coherent evaluation across partial and complete responses, and Preference Consistency, aligning partial sequence assessments with human preferences. To achieve these, we propose $\textbf{SPRM}$, a novel dual-consistency framework integrating score consistency-based and preference consistency-based partial evaluation modules, which leverage the Bradley-Terry model and entropy-based reweighting to predict cumulative rewards and prioritize human-aligned sequences. Extensive experiments on dialogue, summarization, and reasoning tasks demonstrate the effectiveness of SPRM, significantly reducing granularity discrepancies by up to $\textbf{11.7\\%}$ on TL;DR Summarization and achieving a $\textbf{3.6\\%–10.3\\%}$ improvement in GPT-4 evaluation scores across all tasks. Code is publicly available at [this link](https://github.com/xiebin23/SPRM).

Score Consistency Meets Preference Alignment: Dual-Consistency for Partial Reward Modeling

Sequential recommender systems, which leverage historical interactions to deliver targeted recommendations, have been significantly advanced by large language models (LLMs). However, LLM-based generative sequential recommendation often faces two key challenges: the lack of collaborative knowledge and the limited controllability over the generated content. In this paper, we propose a simple Bi-Tuning framework with collaborative information for controllable Large Language Model-based Sequential Recommendation (Laser). Specifically, Bi-Tuning works through incorporating learnable virtual tokens at both the prefix and suffix of the input text, where the prefix tokens enable the adaptation of LLMs with collaborative information, while the suffix token transforms the LLM output into item/user embeddings for similarity comparison, thereby facilitating controllable recommendations. Furthermore, we introduce an MoE-based querying transformer that selectively activates experts to extract relevant information from varying collaborative signals of frozen ID-based recommenders into the prefix, coupled with a multi-task loss function incorporating the MoE load-balancing objective. Finally, a two-phase training strategy is employed to progressively obtain high-quality item and user embeddings through the learnable suffix. Experiments on real-world datasets show that Laser effectively adapts LLMs for sequential recommendation, outperforming state-of-the-art baselines.

Bi-Tuning with Collaborative Information for Controllable LLM-based Sequential Recommendation

Reinforcement learning from human feedback (RLHF) has emerged as the primary method for aligning large language models (LLMs) with human preferences. While it enables LLMs to achieve human-level alignment, it often incurs significant computational and financial costs due to its reliance on training external reward models or human-labeled preferences. In this work, we propose Implicit Preference Optimization (IPO), an alternative approach that leverages generative LLMs as preference classifiers, thereby reducing the dependence on external human feedback or reward models to obtain preferences. We conduct a comprehensive evaluation on the preference classification ability of LLMs using RewardBench, assessing models across different sizes, architectures, and training levels to validate our hypothesis. Furthermore, we investigate the self-improvement capabilities of LLMs by generating multiple responses for a given instruction and employing the model itself as a preference classifier for Direct Preference Optimization (DPO)-based training. Our findings demonstrate that models trained through IPO achieve performance comparable to those utilizing state-of-the-art reward models for obtaining preferences.

IPO: Your Language Model is Secretly a Preference Classifier

*Natural Language to Visualization* (NL2Vis) seeks to convert natural-language descriptions into visual representations of given tables, empowering users to derive insights from large-scale data. Recent advancements in *Large Language Models* (LLMs) show promise in automating code generation to transform tabular data into accessible visualizations. However, they often struggle with complex queries that require reasoning across multiple tables. To address this limitation, we propose a collaborative agent workflow, termed **nvAgent**, for NL2Vis. Specifically, **nvAgent** comprises three agents: a processor agent for database processing and context filtering, a composer agent for planning visualization generation, and a validator agent for code translation and output verification. Comprehensive evaluations on the new VisEval benchmark demonstrate that **nvAgent** consistently surpasses state-of-the-art baselines, achieving a 7.88% improvement in single-table and a 9.23% improvement in multi-table scenarios. Qualitative analyses further highlight that **nvAgent** maintains nearly a 20% performance margin over previous models, underscoring its capacity to produce high-quality visual representations from complex, heterogeneous data sources. All datasets and source code are available at: [https://github.com/geliang0114/nvAgent](https://github.com/geliang0114/nvAgent).

nvAgent: Automated Data Visualization from Natural Language via Collaborative Agent Workflow

We present ZIPA, a family of efficient speech models that advances the state-of-the-art performance of crosslinguistic phone recognition. We first curated IPA PACK++, a large-scale multilingual speech corpus with 17,000+ hours of normalized phone transcriptions and a novel evaluation set capturing unseen languages and sociophonetic variation. ZIPA, including transducer (ZIPA-T) and CTC-based (ZIPA-CR) variants, leverages the efficient Zipformer backbones and outperforms existing phone recognition systems with much fewer parameters. Further scaling via noisy student training on 11,000+ hours of pseudo-labeled multilingual data yields further improvement. While ZIPA achieves strong performance on benchmarks, error analysis reveals persistent limitations in modeling sociophonetic diversity, underscoring challenges for future research.

ZIPA: A family of efficient models for multilingual phone recognition

Large Language Models (LLMs) are increasingly used in working environments for a wide range of tasks, excelling at solving individual problems in isolation. However, are they also able to effectively collaborate over long-term interactions? To investigate this, we introduce MemoryCode, a synthetic multi-session dataset designed to test LLMs' ability to track and execute simple coding instructions amid irrelevant information, simulating a realistic setting. While all the models we tested handle isolated instructions well, even the performance of state-of-the-art models like GPT-4o deteriorates when instructions are spread across sessions. Our analysis suggests this is due to their failure to retrieve and integrate information over long interaction chains. Our results highlight a fundamental limitation of current LLMs, restricting their ability to collaborate effectively in long interactions.

From Tools to Teammates: Evaluating LLMs in Multi-Session Coding Interactions

Argument Mining (AM) involves the automatic identification of argument structure in natural language. Traditional AM methods rely on micro-structural features derived from the internal properties of individual Argumentative Discourse Units (ADUs). However, argument structure is shaped by a macro-structure capturing the functional interdependence among ADUs. This macro-structure consists of segments, where each segment contains ADUs that fulfill specific roles to maintain coherence within the segment (**local coherence**) and across segments (**global coherence**). This paper presents an approach that models macro-structure, capturing both local and global coherence to identify argument structures. Experiments on heterogeneous datasets demonstrate superior performance in both in-dataset and cross-dataset evaluations. The cross-dataset evaluation shows that macro-structure enhances transferability to unseen datasets.

CU-MAM: Coherence-Driven Unified Macro-Structures for Argument Mining

Elasticsearch (ES) is a distributed RESTful search engine optimized for large-scale and long-text search scenarios. Recent research on text-to-Query has explored using large language models (LLMs) to convert user query intent to executable code, making it an increasingly popular research topic. To our knowledge, we are the first to introduce the novel semantic parsing task text-to-ES. To bridge the gap between LLM and ES, in detail, we leverage LLMs and employ domain experts to generate ES query bodies, which are Domain-Specific Language (DSL), along with the corresponding post-processing code to support multi-index ES queries. Consequently, we propose the text-to-ES benchmark that consists of two datasets: Large Elasticsearch Dataset (LED), containing 26,207 text-ES pairs derived from a 224.9GB schema-free database, and ElasticSearch (BirdES)with 10,926 pairs sourced from the Bird dataset on a 33.4GB schema-fixed database. Compared with fourteen advanced LLMs and six code-based LLMs, the model we trained outperformed DeepSeek-R1 by 15.64% on the LED dataset, setting a new state-of-the-art, and achieved 78% of DeepSeek-R1's performance on the BirdES dataset. Additionally, we provide in-depth experimental analyses and suggest future research directions for this task. Our datasets are available at https://huggingface.co/datasets/Barry1915/Text-to-ES.

Downloads

Next from ACL 2025

PVP: An Image Dataset for Personalized Visual Persuasion with Persuasion Strategies, Viewer Characteristics, and Persuasiveness Ratings

Similar lecture

Chinese SafetyQA: A Safety Short-form Factuality Benchmark for Large Language Models