China

Large Language Model (LLM) agents produce rich, multi-step trajectories that interleave observations, internal reasoning, and tool actions. However, most evaluation pipelines focus solely on end-task success, leaving the agent’s decision-making process opaque and poorly understood. We introduce AgentDiagnose, an open-source, modular framework for diagnosing agent trajectories. The present release fully supports the web domain, and AgentDiagnose is architect as an extensible, open platform with compatibility for most agent trajectories. AgentDiagnose consists of (i) an evaluation module that quantifies five core agentic competencies—backtracking &amp; exploration, task decomposition, observation reading, self-verification, and objective quality—and (ii) a visualization module that highlights trajectory semantics through t-SNE action embeddings, interactive word clouds, and state-transition timelines. On a set of 30 manually annotated trajectories, our automatic metrics achieve a mean Pearson correlation of 0.57 with human judgments, rising to 0.78 for task decomposition. Furthermore, filtering the 46k-example NNetNav-Live dataset with AgentDiagnose and fine-tuning a Llama-3.1-8B model on the top 6k trajectories improves WebArena success rates by 0.98, despite using only 13% of the original data. AgentDiagnose thus serves as both a diagnostic lens for agent analysis and a practical tool for curating higher-quality training data. The toolkit and demo are publicly available.

EMNLP 2025

AgentDiagnose: An Open Toolkit for Diagnosing LLM Agent Trajectories

agents

evaluation

The purpose of the AAAI conference series is to promote research in Artificial Intelligence (AI) and foster scientific exchange between researchers, practitioners, scientists, students, and engineers across the entirety of AI and its affiliated disciplines. AAAI-23 is the Thirty-Seventh AAAI Conference on Artificial Intelligence. The theme of this conference is to create collaborative bridges within and beyond AI. Like previous AAAI conferences, AAAI-23 will feature technical paper presentations, special tracks, invited speakers, workshops, tutorials, poster sessions, senior member presentations, competitions, and exhibit programs, and two new activities: a Bridge Program and a Lab Program. Many of these activities are tailored to the theme of bridges and all are selected according to the highest standards, with additional programs for students and young researchers. 
AAAI is providing you with a conference planner, which you can use to help organize your itinerary of activities. This includes talks to attend in person, talks to attend remotely, breaks with colleagues and your site seeing activities. To access this conference planner, please go to [https://aaai-2023.takemobi.io](https://aaai-2023.takemobi.io).

In order to access this site, you need to register. If you haven't already, please register [here](https://aaai.org/Conferences/AAAI-23/registration/).


AAAI 2023

The purpose of the AAAI conference series is to promote research in Artificial Intelligence (AI) and foster scientific exchange between researchers, practitioners, scientists, students, and engineers across the entirety of AI and its affiliated disciplines.

Large Language Model (LLM) agents produce rich, multi-step trajectories that interleave observations, internal reasoning, and tool actions. However, most evaluation pipelines focus solely on end-task success, leaving the agent’s decision-making process opaque and poorly understood. We introduce AgentDiagnose, an open-source, modular framework for diagnosing agent trajectories. The present release fully supports the web domain, and AgentDiagnose is architect as an extensible, open platform with compatibility for most agent trajectories. AgentDiagnose consists of (i) an evaluation module that quantifies five core agentic competencies—backtracking & exploration, task decomposition, observation reading, self-verification, and objective quality—and (ii) a visualization module that highlights trajectory semantics through t-SNE action embeddings, interactive word clouds, and state-transition timelines. On a set of 30 manually annotated trajectories, our automatic metrics achieve a mean Pearson correlation of 0.57 with human judgments, rising to 0.78 for task decomposition. Furthermore, filtering the 46k-example NNetNav-Live dataset with AgentDiagnose and fine-tuning a Llama-3.1-8B model on the top 6k trajectories improves WebArena success rates by 0.98, despite using only 13% of the original data. AgentDiagnose thus serves as both a diagnostic lens for agent analysis and a practical tool for curating higher-quality training data. The toolkit and demo are publicly available.

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.

Despite Bengali being the sixth most spoken language in the world, handwritten text recognition (HTR) systems for Bengali remain severely underdeveloped. The complexity of Bengali script—featuring conjuncts, diacritics, and highly variable handwriting styles—combined with a scarcity of annotated datasets makes this task particularly challenging. We present **GraDeT-HTR**, a resource-efficient Bengali handwritten text recognition system based on a decoder-only transformer architecture. To address the unique challenges of Bengali script, we integrate a grapheme-based tokenizer and demonstrate that it significantly improves recognition accuracy compared to conventional subword tokenizers. Our model is pretrained on large-scale synthetic data and fine-tuned on real human-annotated samples, achieving state-of-the-art performance on multiple benchmark datasets.

GraDeT-HTR: A Resource-Efficient Bengali Handwritten Text Recognition System utilizing Grapheme-based Tokenizer and Decoder-only Transformer

The ARR Responsible NLP Research checklist website states that the "checklist is designed to encourage best practices for responsible research, addressing issues of research ethics, societal impact and reproducibility." Answering the questions is an opportunity for authors to reflect on their work and make sure any shared scientific assets follow best practices. Ideally, considering a checklist before submission can favorably impact the writing of a research paper. However, previous research has shown that self-reported checklist responses don’t always accurately represent papers. In this work, we introduce ConfReady, a retrieval-augmented generation (RAG) application that can be used to empower authors to reflect on their work and assist authors with conference checklists. To evaluate checklist assistants, we curate a dataset of 1975 ACL checklist responses, analyze problems in human answers, and benchmark RAG and Large Language Model (LM) based systems on a evaluation subset. Our code is released under the AGPL-3.0 license on GitHub, with documentation covering the user interface and Python package.

ConfReady: A RAG based Assistant and Dataset for Conference Checklist Responses

The rapid advancement of generative AI has democratized access to powerful tools such as Text-to-Image (T2I) models. However, to generate high-quality images, users must still craft detailed prompts specifying scene, style, and context—often through multiple rounds of refinement. We propose PromptSculptor, a novel multi-agent framework that automates this iterative prompt optimization process. Our system decomposes the task into four specialized agents that work collaboratively to transform a short, vague user prompt into a comprehensive, refined prompt. By leveraging Chain-of-Thought (CoT) reasoning, our framework effectively infers hidden context and enriches scene and background details. To iteratively refine the prompt, a self-evaluation agent aligns the modified prompt with the original input, while a feedback-tuning agent incorporates user feedback for further refinement. Experimental results demonstrate that PromptSculptor significantly enhances output quality and reduces the number of iterations needed for user satisfaction. Moreover, its model-agnostic design allows seamless integration with various T2I models, paving the way for industrial applications.

PromptSculptor: Multi-Agent Based Text-to-Image Prompt Optimization

Accurate clinical coding is essential for healthcare documentation, billing, and decision-making. While prior work shows that off-the-shelf LLMs struggle with this task, evaluations based on exact match metrics often overlook errors where predicted codes are hierarchically close but incorrect. Our analysis reveals that such hierarchical misalignments account for a substantial portion of LLM failures. We show that lightweight interventions, including prompt engineering and small-scale fine-tuning, can improve accuracy without the computational overhead of search-based methods. To address hierarchically near-miss errors, we introduce clinical code verification as both a standalone task and a pipeline component. To mitigate the limitations in existing datasets, such as incomplete evidence and inpatient bias in MIMIC, we release an expert double-annotated benchmark of outpatient clinical notes with ICD-10 codes. Our results highlight verification as an effective and reliable step toward improving LLM-based medical coding.

Toward Reliable Clinical Coding with Language Models: Verification and Lightweight Adaptation

We present a robust framework for deploying domain-specific language agents that can query industrial sensor data using natural language. Grounded in the Reasoning and Acting (ReAct) paradigm, our system introduces three key innovations: (1) integration of the \texttt{Self-Ask} method for compositional, multi-hop reasoning; (2) a multi-agent architecture with \texttt{Review}, \texttt{Reflect} and \texttt{Distillation} components to improve reliability and fault tolerance; and (3) a long-context prompting strategy leveraging curated in-context examples, which we call \textit{Tiny Trajectory Store}, eliminating the need for fine-tuning. We apply our method to Industry 4.0 scenarios, where agents query SCADA systems (e.g., SkySpark) using questions such as, “How much power did B002 AHU 2-1-1 use on 6/14/16 at the POKMAIN site?” To enable systematic evaluation, we introduce \textbf{IoTBench}, a benchmark of 400+ tasks across five industrial sites. Our experiments show that ReAct-style agents enhanced with long-context reasoning (\texttt{ReActXen}) significantly outperform standard prompting baselines across multiple LLMs including smaller models. This work repositions NLP agents as practical interfaces for industrial automation, bridging natural language understanding and sensor-driven environments.

ReAct Meets Industrial IoT: Language Agents for Data Access

Trends on microblogs often transcend linguistic boundaries, evolving into global phenomena with significant societal and economic impact. We address the novel task of predicting "cross-lingual trends" on microblog that originate in one language and subsequently become popular in others. While crucial for global monitoring and marketing, this area has been under-explored due to the challenge of cross-lingual trend identification. We introduce a methodology to overcome this by automatically constructing a dataset using Wikipedia's inter-language links to reconcile trend names. We propose a prediction model that leverages a rich feature set, including not only temporal frequency but also microblog content and external knowledge signals from Wikipedia (e.g., article content, pageviews). Our approach achieves higher accuracy than existing trend prediction methods and LLM-based approaches, enabling effective early detection of cross-lingual trends.

Predicting Cross-lingual Trends in Microblogs

The increasing complexity of modern driving systems demands efficient collection and analysis of specific driving scenarios that are crucial for system development and validation. Current approaches either rely on massive data collection followed by manual filtering, or rigid threshold-based recording systems that often miss important edge cases. In this paper, we present Distributed Adaptive Scene Recognition (DASR), a novel multi-agent cloud-edge framework for language-guided scene detection in connected vehicles. Our system leverages the complementary strengths of cloud-based large language models and edge-deployed vision language models to intelligently identify and preserve relevant driving scenarios while optimizing limited on-vehicle buffer storage. The cloud-based LLM serves as an intelligent coordinator that analyzes developer prompts to determine which specialized tools and sensor data streams should be incorporated, while the edge-deployed VLM efficiently processes video streams in real time to make relevant decisions. Extensive experiments across multiple driving datasets demonstrate that our framework achieves superior performance compared to larger baseline models, with exceptional performance on complex driving tasks requiring sophisticated reasoning. DASR also shows strong generalization capabilities on out-of-distribution datasets and significantly reduces storage requirements (28.73 %) compared to baseline methods.

DASR: Distributed Adaptive Scene Recognition - A Multi-Agent Cloud-Edge Framework for Language-Guided Scene Detection

For an e-commerce domain, the customer
address is the single most important piece
of customer data for ensuring accurate
and reliable deliveries. In this two-part
study, we first outline the construction of
a language model to assist customers with
address standardization and in the latter
part, we detail a novel Pareto-ensemble
multi-task prediction algorithm that derives critical insights from customer addresses to minimize operational losses 
arising from a given geographical area. Finally, 
we demonstrate the potential benefits of
the proposed address intelligence system
for a large e-commerce domain through
large scale experiments on a commercial
system.

An Address Intelligence Framework for E-commerce Deliveries

Adapting language models to learn continuously from data streams while retaining previous knowledge is a key challenge in artificial intelligence (AI), particularly in lifelong language learning. Existing distillation methods are based on offline techniques, limiting their ability to update in real-time and adapt to dynamic environments. To address this, we propose online dynamic mutual distillation - a novel framework that enables continuous mutual learning from task streams without relying on domain-specific teachers. To our knowledge, this is the first application of mutual learning in lifelong language learning, providing dynamic knowledge transfer without domain-specific teachers. Moreover, our extensive experiments demonstrate that the proposed method reduces catastrophic forgetting, while improving task performance on various benchmark datasets making it suitable for real-world, dynamic natural language processing (NLP) applications such as adaptive chatbots and personalized language systems. We will make our code publicly available upon acceptance.

L4: Mutual Learning Helps Lifelong Language Learning

The rapid advancements in Large Language Models (LLMs) have enabled their adoption in real-world industrial scenarios for various natural language processing tasks. However, the high inference cost of large-scale LLMs makes their deployment impractical, necessitating the use of smaller models. Despite their efficiency, smaller LLMs lack robust zero-shot instruction-following capabilities across diverse domains, limiting their adaptability to dynamic user requirements. Traditional fine-tuning approaches exacerbate this issue by inducing catastrophic forgetting, reducing the model’s generalization ability for unseen tasks. In this paper, we propose Domain Adaptive Continual Instruction Pre-Training via Reading Comprehension (DACIP-RC), a continual pre-training technique that enhances smaller LLMs' domain adaptability for business conversational tasks. Unlike conventional pre-training approaches that rely on next-token prediction, DACIP-RC generates diverse task instructions and responses via reading comprehension on conversation transcripts, enabling better instruction generalization. Our empirical evaluations demonstrate that DACIP-RC significantly improves zero-shot generalization across a wide range of business conversational tasks, including meeting summarization, action item generation, and call purpose identification. To the best of our knowledge, this is the first work to apply instruction pre-training on business conversational data, providing insights into how industries can leverage proprietary datasets for domain adaptation.

Downloads

Next from EMNLP 2025

GraDeT-HTR: A Resource-Efficient Bengali Handwritten Text Recognition System utilizing Grapheme-based Tokenizer and Decoder-only Transformer

Similar lecture

Learning From Good Trajectories in Offline Multi-Agent Reinforcement Learning

Stay up to date with the latest Underline news!

PRESENTATIONS

CONFERENCES

COMPANY

RESOURCES