China

LLMs are increasingly applied in the legal domain in tasks such as summarizing legal texts and providing basic legal advice. Yet, their capacity to draft full judicial analyses in U.S. court opinions is still largely uncharted, such as generating entire judicial reasoning sections in U.S. court decisions, remain under-explored. Given the continued adoption of LLMs and the significance of law to society at large, measurement of LLM&#39;s legal reasoning capabilities is a pressing task. We propose CourtReasoner, a novel expert-annotated judicial reasoning benchmark for evaluating LLM agents&#39; capabilities in complex legal reasoning. Sourcing U.S. court opinions, we construct benchmarks that measure the LLMs ability to construct goal-oriented legal reasoning. CourtReasoner measured the agent&#39;s ability to argue both ways in a legal dispute, rather than simple Q/A. Our results show that more than 60% of frontier model outputs contain invalid arguments and more than 53% of frontier model produced irrelevant citations when conducting complex legal reasoning. We also introduce a meta-evaluation benchmark to provide insights into the capabilities of LLMs as evaluators of legal reasoning. We will release our data, code and full annotation guidelines publicly for future research.

EMNLP 2025

CourtReasoner: Can LLM Agents Reason Like Judges?

LLMs are increasingly applied in the legal domain in tasks such as summarizing legal texts and providing basic legal advice. Yet, their capacity to draft full judicial analyses in U.S. court opinions is still largely uncharted, such as generating entire judicial reasoning sections in U.S. court decisions, remain under-explored. Given the continued adoption of LLMs and the significance of law to society at large, measurement of LLM's legal reasoning capabilities is a pressing task. We propose CourtReasoner, a novel expert-annotated judicial reasoning benchmark for evaluating LLM agents' capabilities in complex legal reasoning. Sourcing U.S. court opinions, we construct benchmarks that measure the LLMs ability to construct goal-oriented legal reasoning. CourtReasoner measured the agent's ability to argue both ways in a legal dispute, rather than simple Q/A. Our results show that more than 60% of frontier model outputs contain invalid arguments and more than 53% of frontier model produced irrelevant citations when conducting complex legal reasoning. We also introduce a meta-evaluation benchmark to provide insights into the capabilities of LLMs as evaluators of legal reasoning. We will release our data, code and full annotation guidelines publicly for future research.

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."<br>

*Message from the General Chair, Dirk Hovy*

[**Link to Conference Handbook**](https://drive.google.com/file/d/1johU5QqVVYO4RfH7QcIORr7qrVBdzdwC/view?usp=sharing)





<br>

Celebrate 30 Years of EMNLP! 
EMNLP 2025 will be held in Suzhou, China from November 5th to November 9th, 2025.

Regulatory agencies often operate with limited resources and rely on tips from the public to identify potential violations. However, processing these tips at scale presents significant operational challenges, as agencies must correctly identify and route relevant tips to the appropriate enforcement divisions. Through a case study, we demonstrate how advances in large language models can be utilized to support overburdened agencies with limited capacities. In partnership with the U.S. Environmental Protection Agency, we leverage previously unstudied citizen tips data from their “Report a Violation” system to develop an LLM-assisted pipeline for tip routing. Our approach filters out 75.2% of irrelevant tips and increases overall routing accuracy from 32.8% to 81.2% compared to the current routing system. At a time of increased focus on government efficiencies, our approach provides a constructive path forward by using technology to empower civil servants.

Not Your Typical Government Tipline: LLM-Assisted Routing of Environmental Protection Agency Citizen Tips

Transformer-based speech language models (SLMs) have significantly improved neural speech recognition and understanding. While existing research has examined how well SLMs encode shallow acoustic and phonetic features, the extent to which SLMs encode nuanced syntactic and conceptual features remains unclear. By drawing parallels with linguistic competence assessments for large language models, this study is the first to systematically evaluate the presence of contextual syntactic and semantic features across SLMs for self-supervised learning (S3M), automatic speech recognition (ASR), speech compression (codec), and as the encoder for auditory large language models (AudioLLMs). Through minimal pair designs and diagnostic feature analysis across 71 tasks spanning diverse linguistic levels, our layer-wise and time-resolved analysis uncovers that 1) all speech encode grammatical features more robustly than conceptual ones. 2) Despite never seeing text, S3M match or surpass ASR encoders on every linguistic level, demonstrating that rich grammatical and even conceptual knowledge can arise purely from audio. 3) S3M representations peak mid-network and then crash in the final layers, whereas ASR and AudioLLM encoders maintain or improve, reflecting how pre-training objectives reshape late-layer content. 4) Temporal probing further shows that S3Ms encode grammatical cues 500 ms before a word begins, whereas AudioLLMs distribute evidence more evenly—indicating that objectives shape not only where but also when linguistic information is most salient. Together, these findings establish the first large-scale map of contextual syntax and semantics in speech models and highlight both the promise and the limits of current SLM training paradigms.

Layer-wise Minimal Pair Probing Reveals Contextual Grammatical-Conceptual Hierarchy in Speech Representations

The rise of Multi-Agent Systems (MAS) in Artificial Intelligence (AI), especially integrated with Large Language Models (LLMs), has greatly facilitated the resolution of complex tasks. However, current systems are still facing challenges of inter-agent communication, coordination, and interaction with heterogeneous tools and resources. Most recently, the Model Context Protocol (MCP) by Anthropic and Agent-to-Agent (A2A) communication protocol by Google have been introduced, and to the best of our knowledge, very few applications exist where both protocols are employed within a single MAS framework. We present a pilot study of AgentMaster, a novel modular multi-protocol MAS framework with self-implemented A2A and MCP, enabling dynamic coordination, flexible communication, and rapid development with faster iteration. Through a unified conversational interface, the system supports natural language interaction without prior technical expertise and responds to multimodal queries for tasks including information retrieval, question answering, and image analysis. The experiments are validated through both human evaluation and quantitative metrics, including BERTScore F1 (96.3%) and LLM-as-a-Judge G-Eval (87.1%). These results demonstrate robust inter-agent coordination, query decomposition, dynamic routing, and domain-specific relevant responses. Overall, our proposed framework contributes to the potential capabilities of domain-specific, cooperative, and scalable conversational AI powered by MAS.

AgentMaster: A Multi-Agent Conversational Framework Using A2A and MCP Protocols for Multimodal Information Retrieval and Analysis

The learning process for medical residents presents significant challenges, demanding both the ability to interpret complex case reports and the rapid acquisition of accurate medical knowledge from reliable sources. Residents typically study case reports and engage in discussions with peers and mentors, but finding relevant educational materials and evidence to support their learning from these cases is often time-consuming and challenging. To address this, we introduce MedTutor, a novel system designed to augment resident training by automatically generating evidence-based educational content and multiple-choice questions from clinical case reports. MedTutor leverages a Retrieval-Augmented Generation (RAG) pipeline that takes clinical case reports as input and produces targeted educational materials. The system's architecture features a hybrid retrieval mechanism that synergistically queries a local knowledge base of medical textbooks and academic literature (using PubMed, Semantic Scholar APIs) for latest related research, ensuring the generated content is both foundationally sound and current. The retrieved evidence is filtered and ordered using a state-of-the-art reranking model and then an LLM generates the final long-form output describing the main educational content regarding the case-report. We conduct a rigorous evaluation of the system. First, two radiologists assessed the quality of outputs, finding them to be of high clinical and educational value. Second, we perform a large-scale evaluation using an LLM-as-a Judge to understand if LLMs can be used to evaluate the output of the system. Our analysis using correlation of LLMs with human expert judgments reveals a moderate alignment and highlights the continued necessity of expert oversight.

MedTutor: A Retrieval-Augmented LLM System for Case-Based Medical Education

Recent advancements in large language models (LLMs) have significantly transformed the medical systems. However, their potential within specialized domains such as nursing remains largely underexplored. In this work, we introduce NurseLLM, the first nursing-specialized LLM tailored for nursing multiple choice question-answering (MCQ) tasks. We develop a multi-stage data generation pipeline to build the first large scale nursing MCQ dataset to train LLMs on a broad spectrum of nursing topics. We further introduce multiple nursing benchmarks to enable rigorous evaluation. Our extensive experiments demonstrate that NurseLLM outperforms SoTA general-purpose and medical-specialized LLMs of comparable size on different benchmarks, underscoring the importance of a specialized LLM for the nursing domain. Finally, we explore the role of reasoning and multi-agent collaboration systems in nursing, highlighting their promise for future research and applications.

NurseLLM: The First Specialized Language Model for Nursing

In modern industry systems like multi-turn chat agents, Text-to-SQL technology bridges natural language (NL) questions and database (DB) querying, and conversion of tabular DB results into NL representations (NLRs) enables the chat format. Currently, NLR generation is typically handled by large language models (LLMs), but information loss or errors in presenting tabular results in NL remains largely unexplored.
This paper introduces a novel evaluation method - Combo-Eval - for judgment of LLM-generated NLRs that combines the benefits of multiple existing methods, optimizing evaluation fidelity and achieving a significant reduction in LLM calls by 25-61%. Accompanying our method is NLR-BIRD, the first dedicated dataset for NLR benchmarking. Through human evaluations, we demonstrate the superior alignment of Combo-Eval with human judgments, applicable across scenarios with and without ground truth references.

Can LLMs Narrate Tabular Data? An Evaluation Framework for Natural Language Representations of Text-to-SQL System Outputs

Accurate intent classification is critical for efficient routing in customer service, ensuring customers are connected with the most suitable agents while reducing handling times and operational costs. However, as companies expand their product lines, intent classification faces scalability challenges due to the increasing number of intents and variations in taxonomy across different verticals. In this paper, we introduce REIC, a Retrieval-augmented generation Enhanced Intent Classification approach, which addresses these challenges effectively. REIC leverages retrieval-augmented generation (RAG) to dynamically incorporate relevant knowledge, enabling precise classification without the need for frequent retraining. Through extensive experiments on real-world datasets, we demonstrate that REIC outperforms traditional fine-tuning, zero-shot, and few-shot methods in large-scale customer service settings. Our results highlight its effectiveness in both in-domain and out-of-domain scenarios, demonstrating its potential for real-world deployment in adaptive and large-scale intent classification systems.

REIC: RAG-Enhanced Intent Classification at Scale

Short video platforms are evolving rapidly, making the identification of inappropriate content increasingly critical.
Existing approaches typically train separate and small classification models for each type of issue, which requires extensive human-labeled data and lacks cross-issue generalization.
We propose a reasoning-enhanced multimodal large language model (MLLM) pretraining paradigm for unified inappropriate content detection. To address the distribution gap between short video content and the original pretraining data of MLLMs, as well as the complex issue definitions, we introduce three targeted pretraining tasks:
(1) \textit{Caption}, to enhance the MLLM's perception of video details;
(2) \textit{Visual Question Answering (VQA)}, to deepen the MLLM's understanding of issue definitions and annotation guidelines;
(3) \textit{Chain-of-Thought (CoT)}, to enhance the MLLM's reasoning capability.
Experimental results show that our pretraining approach significantly improves the MLLM's performance in both zero-shot and supervised fine-tuning (SFT) settings.
In addition, our pretrained model demonstrates strong generalization capabilities to emergent, previously unseen issues.

Reasoning-Enhanced Domain-Adaptive Pretraining of Multimodal Large Language Models for Short Video Content Moderation

Natural language interfaces (NLIs) democratize data analytics by enabling non-technical users to query relational databases via Text-to-SQL systems. While large language models (LLMs) have achieved state-of-the-art accuracy on benchmarks like Spider and BIRD, two critical challenges persist for real-time deployment: (1) inference latency due to sequential autoregressive decoding (e.g., 14.23–22.77 seconds per query for Qwen2.5-Coder-32B and Llama-70B on BIRD (Minidev)), and (2) schema hallucinations (e.g., invalid column references like customer_ids instead of cust_id). To address these, we propose Tree-Guided Token Decoding (TTD-SQL), a lightweight framework that integrates SQL grammar and database schema constraints into the decoding process without modifying the underlying LLM. TTD precomputes token-level decision trees over SQL keywords, table names, and column identifiers, enabling deterministic "auto-fill" transitions for uniquely determined tokens (e.g., "Singer_" → "ID") while retaining flexibility for unconstrained reasoning. Across five LLMs (CodeLlama, Phi-4, Qwen2.5, Granite, Llama-70B), TTD achieves up to 19.96% token-rate speedups by eliminating redundant forward passes (e.g., CodeLlama: 8.97→10.76 tokens/s on Spider) and reduces schema hallucinations by +17.7% in executable-SQL rates (e.g., CodeLlama on BIRD). By bridging rigid parser-based methods and flexible LLM generation, TTD offers a practical path toward reliable, high-performance SQL generation in both public benchmarks and enterprise settings.

TTD-SQL: Tree-Guided Token Decoding for Efficient and Schema-Aware SQL Generation

AI-generated clinical notes are increasingly used in healthcare, but evaluating their quality remains a challenge due to high subjectivity and limited scalability of expert review. Existing automated metrics often fail to align with real-world physician preferences. To address this, we propose a pipeline that systematically distills real user feedback into structured checklists for note evaluation. These checklists are designed to be interpretable, grounded in human feedback, and enforceable by LLM-based evaluators. Using deidentified data from over 21,000 clinical encounters (prepared in accordance with the HIPAA safe harbor standard) from a deployed AI medical scribe system, we show that our feedback-derived checklist outperforms baseline approaches in our offline evaluations in coverage, diversity, and predictive power for human ratings. Extensive experiments confirm the checklist's robustness to quality-degrading perturbations, significant alignment with clinician preferences, and practical value as an evaluation methodology. In offline research settings, our checklist offers a practical tool for flagging notes that may fall short of our defined quality standards.

Downloads

Next from EMNLP 2025

Not Your Typical Government Tipline: LLM-Assisted Routing of Environmental Protection Agency Citizen Tips

Stay up to date with the latest Underline news!

PRESENTATIONS

CONFERENCES

COMPANY

RESOURCES