United States

Conversational AI (ConvAI) agents increasingly maintain structured memory to support long-term, task-oriented interactions. In-context memory approaches append the growing history to the model input, which scales poorly under context-window limits. RAG-based methods retrieve request-relevant information, but most assume flat memory collections and ignore structure. We propose **Semantic XPath**, a **tree-structured memory module** to access and update structured conversational memory. **Semantic XPath** improves performance over flat-RAG baselines by **176.7%** while using only **9.1%** of the tokens required by in-context memory. We also introduce **SemanticXPath Chat**, an end-to-end ConvAI demo system that visualizes the structured memory and query execution details. Overall, this paper demonstrates a candidate for the next generation of long-term, task-oriented ConvAI systems built on structured memory.

ACL 2026

Semantic XPath: Structured Agentic Memory Access for Conversational AI

It is my great pleasure and honor to welcome you to the 64th Annual Meeting of the Association for Computational Linguistics (ACL 2026), held in San Diego, California, USA, July 2–7, 2026. ACL 2026 continues our fields 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.

It is exciting to see the many developments in the field reflected at our conference, such as the advancement of large language models (born out of our community) into many application directions. This is not without challenges, partly due to unsolved problems, and partly due to misuse of the technology. All this is reflected well at ACL 2026. The conference also reflects that the field is not only about advancing technology, but also the science of fundamentally understanding language and how it is use by many different speaker communities. <br>
[Full Message from General Chair Philipp Koehn](https://drive.google.com/file/d/1QiJFOmMjvDgsDo3i1XzlQnImxe3RU_Cy/view?usp=sharing) <br><br><br>
<html><button style="display: inline-flex; align-items: center; justify-content: center; white-space: nowrap; border-radius: 9999px; font-weight: bold; background: #7c3aed; color: white; font-family: 'Space Grotesk', sans-serif; height: 40px; font-size: 16px; padding: 0 20px; border: none; cursor: pointer" onclick="window.open('https://drive.google.com/file/d/1QiJFOmMjvDgsDo3i1XzlQnImxe3RU_Cy/view?usp=sharing')">Conference Handbook</button></html>
<br><br><br><br>

To access this event page, you need to log in with the **email address you registered with**. <br>Access credentials will be sent to your email from Underline -  subject line "Welcome to ACL 2026". Please be sure to check your spam email folder if you do not see an email confirmation right away.

To access this event page, you are required to register.
Please complete your registration to continue.

We recommend reading [**the registration information**](https://2026.aclweb.org/registration/) first.

**Online Registration Form**: https://acl.swoogo.com/acl2026

Registration Required

The 64th Annual Meeting of the Association for Computational Linguistics (ACL 2026) will take place in San Diego, California, from July 2 - 7, 2026.

demo

Annotated data remains essential for training and evaluating NLP systems. Large language models have increased the scope of what type of data is needed, creating new needs for multimodal and agentic system data. Here, we introduce Potato 2.0, a major update to our open source annotation platform designed for easy deployment, customization, and fully reproducible and sharable annotation designs. Potato offers broad support for many types of annotations in NLP, including 15 different types of annotation tasks, support for text, audio, image, and video modalities---or mixtures thereof,---and the support for labeling agentic system outputs or live interaction with agents. Potato also includes multiple AI-assistance features to help annotators more easily label data. Finally, Potato introduces a new agentic AI-in-the-loop workflow where a single human annotator collaborates with an LLM through iterative prompt refinement, uncertainty-driven instance selection, and progressive autonomy---enabling efficient dataset creation without a large annotation team.

Potato 2.0: A Comprehensive Annotation Platform with AI-in-the-Loop Support

We present mllm-shap, an open-source
Python platform for researchers and ML practitioners that extends Shapley value (SV)
explainability from text-only large language
models to multimodal LLMs (MLLMs) that
jointly process text and audio. Building
on the token-level SV framework introduced
by TokenSHAP, mllm-shap addresses three
challenges absent in the text-only setting:
(1) modality-aware coalition masking that
handles the coexistence of text tokens and
dense audio encoder frames within a single input, (2) multi-turn conversation tracking with
per-token role and modality metadata, and
(3) audio token grouping via phonetic alignment that reduces the coalition space by 10–50 times. The platform ships as a pip-installable
package implementing five SV estimation strategies – including a Complementary Contributions estimator with Neyman-optimal allocation that outperforms Monte Carlo baselines – together with an interactive web GUI
for real-time attribution visualization. To
our knowledge, mllm-shap is the first publicly available framework for complete, reproducible SV-based explainability of text-audio
MLLMs. The package is MIT-licensed with
full source code on GitHub and a demonstration video included as supplementary material.

mllm-shap: A Shapley Value Explainability Platform for Text-Audio Multimodal Large Language Models

Understanding the emotions that individuals in crisis express is a clinically relevant goal. Here, we introduce an automated method for extracting present and past personal emotion expressions from text-based crisis conversations, enabling nuanced analyses of how these emotional profiles vary by age. We develop a three-tier emotion taxonomy and leverage both real conversation data and synthetic sentences to train a transformer-based model that captures contextual distinctions between true personal emotion expressions and other mentions. Our RoBERTa-based classifier outperforms both a regex baseline and a model trained only on real conversation data, achieving an F1 score of 0.856. Subsequent analysis of 338,924 crisis conversations shows that age is correlated with distinct patterns in emotional expressions. These findings underscore the clinical value of age-sensitive emotion analysis and constitute an initial step toward characterizing lexical variations across demographic groups.

Extraction of Texters' Explicit Emotion Expressions in Crisis Conversations

Event annotation is important for identifying, monitoring, and understanding sociological trends. Although expert annotators set the gold standard, they are expensive and inefficient. While state-of-the-art NLP models are an attractive alternative, they are often evaluated on standalone subtasks rather than entire workflows. Thus, we evaluate a holistic workflow that summarizes news with event coreference resolution and argument extraction in three modes: AI-only, AI assistance, and human only. Although AI's recall is seven times higher than the tf-idf baseline at coreference resolution, it is far from replacing experts. However, experts adopt AI-extracted arguments 60% of the time, reducing extraction time by 25%. Our code and data are in https://github.com/Obertura777/gtd-data.

Large Language Models Are Effective Human Annotation Assistants, But Not Good Independent Annotators

Large Language Models (LLMs) have fundamentally transformed natural language processing (NLP), demonstrating remarkable capabilities across a wide spectrum of tasks. However, when applied to instruction-based text editing, LLMs continue to exhibit some limitations. Different from free-form generation, instruction-based editing requires precise, targeted modifications that respect two essential properties: faithfully implementing the specific instruction and local fidelity. Existing approaches often overlook these properties, treating editing as a generic text generation problem. As a result, they either over-edit or fail to apply modifications consistently. To address this gap, we propose HyperEdit, a framework that adaptively processes each editing request to best align with it. To achieve this, HyperEdit generates request-specific dynamic weights that guide the editing process. The computational overhead of producing these weights is minimized through a carefully designed hypernetwork. With this design, HyperEdit achieves a relatively 9% improvement over the state-of-the-art editing model.

HyperEdit: Unlocking Instruction-based Text Editing in LLMs via Hypernetworks

Evaluating how large language models (LLMs) capture the grammatical structure of low-resource languages remains underexplored. This paper presents the Urdu Benchmark of Linguistic Minimal Pairs (UrBLiMP)—a diagnostic suite of 5,696 minimal pairs that contrast grammatical acceptability across ten core syntactic and morpho-syntactic phenomena in Urdu. The dataset is constructed from the Urdu Treebank and diverse text corpora, and human validation achieves a 96.1\% inter-annotator agreement, confirming its reliability. We evaluate twenty one multilingual LLMs, including LLaMA-3-70B and Gemma-3-27B-PT, and additionally assess the proprietary GPT-4o model using grammar-prompting techniques. GPT-4o (grammar-prompted) attains the highest average accuracy (97.4\%), reaching near-human performance on regular phenomena such as aspect agreement and ergativity. However, all models continue to struggle with flexible syntactic patterns like word-order variation and long-distance subject–verb agreement. UrBLiMP provides the first controlled evaluation framework for probing morpho-syntactic competence in Urdu and highlights both the progress and remaining challenges of multilingual and proprietary LLMs in low-resource settings.

 UrBLiMP: A Benchmark for Evaluating the Linguistic Competence of Large Language Models in Urdu

Retrieval over visually rich documents is essential for tasks such as legal discovery, scientific search, and enterprise knowledge management. Existing approaches fall into two paradigms: single-vector retrieval, which is efficient but coarse, and multi-vector retrieval, which is accurate but computationally expensive. To address this trade-off, we propose HEAVEN, a plug-and-play two-stage hybrid-vector framework. In the first stage, HEAVEN efficiently retrieves candidate pages using a single-vector method over Visually-Summarized Pages (VS-Pages), which assemble representative visual layouts from multiple pages. In the second stage, it reranks candidates with a multi-vector method while filtering query tokens by linguistic importance to reduce redundant computations. To evaluate retrieval systems under realistic conditions, we also introduce ViMDoc, a benchmark for visually rich, multi-document, and long-document retrieval. Across four benchmarks, HEAVEN attains 99.87% of the Recall@1 performance of multi-vector models on average while reducing per-query computation by 99.82%, achieving efficiency and accuracy. Our code and datasets are available at: https://github.com/juyeonnn/HEAVEN

Hybrid-Vector Retrieval for Visually Rich Documents: Combining Single-Vector Efficiency and Multi-Vector Accuracy

Prompt injection attacks manipulate large language models (LLMs) by misleading them to deviate from the original input instructions and execute maliciously injected instructions, because of their instruction-following capabilities and inability to distinguish between the original input instructions and maliciously injected instructions. Currently, various prompt injection defense methods have been proposed, including prompt-engineering-based approaches and fine-tuning methods. Most of these methods instruct the model to follow the original input instructions, suppressing its inherent tendencies to follow the injected instructions. However, experimental results reveal that suppressing the model's instruction-following tendencies is challenging. After analyzing successful attack cases, we find that the LLMs can correctly reference the instructions they are executing in some cases. Motivated by this finding, we propose a defense method that leverages LLMs’ instruction-following abilities rather than suppressing them. Our approach prompts LLMs to generate responses that include both the answers and their corresponding instruction references. Based on these references, we filter out answers whose references are not to the original input instructions. We conduct comprehensive experiments to evaluate the effectiveness of our proposed method. The results show that our approach outperforms prompt-engineering-based baselines and is comparable to fine-tuning methods, reducing the ASR to nearly 0% in some scenarios. Moreover, our approach has minimal impact on overall utility.

Robustness via Referencing: Defending against Prompt Injection Attacks by Referencing the Executed Instruction

High-quality datasets are critical for training and evaluating reliable NLP models. In tasks like natural language inference (NLI), human label variation (HLV) arises when multiple labels are valid for the same instance, making it difficult to separate annotation errors from plausible variation. An earlier framework, VariErr (Weber-Genzel et al., 2024), asks multiple annotators to explain their label decisions in the first round and flags errors through validity judgments in the second round. However, conducting two rounds of manual annotation is costly and may limit the coverage of plausible labels or explanations. Our study proposes a new framework, EVADE, for generating and validating explanations to detect errors using large language models (LLMs). We perform a comprehensive analysis comparing human- and LLM-detected errors for NLI across distribution comparison, validation overlap, and impact on model fine-tuning. Our experiments demonstrate that LLM validation refines generated explanation distributions to more closely align with human annotations, and that removing LLM-detected errors from training data yields improvements in fine-tuning performance than removing errors identified by human annotators. This highlights the potential to scale error detection, reducing human effort while improving dataset quality under label variation.

EVADE: LLM-Based Explanation Generation and Validation for Error Detection in NLI

Large language models (LLMs) require continual updates to rectify outdated or erroneous knowledge. Model editing has emerged as a compelling paradigm for introducing targeted modifications without the computational burden of full retraining. Existing approaches are mainly based on a locate-then-edit framework. However, in sequential editing contexts, where multiple updates are applied over time, they exhibit significant limitations and suffer from catastrophic interference, i.e., new edits compromise previously integrated updates and degrade preserved knowledge. To address these challenges, we introduce EvoEdit, a novel editing strategy that mitigates catastrophic interference through sequential null-space alignment, enabling stable and efficient model editing. By performing sequential null-space alignment for each incoming edit, EvoEdit preserves both original and previously modified knowledge representations and maintains output invariance on preserved knowledge even across long edit sequences, effectively mitigating interference. Evaluations on real-world sequential knowledge-editing benchmarks show that EvoEdit achieves better or comparable performance than prior state-of-the-art locate-then-edit techniques, with up to 3.53× speedup. Overall, these results underscore the necessity of developing more principled approaches for designing LLMs in dynamically evolving information settings, while providing a simple yet effective solution with strong theoretical guarantees.

Downloads

Next from ACL 2026

Potato 2.0: A Comprehensive Annotation Platform with AI-in-the-Loop Support

Similar lecture

From Recall to Forgetting: Benchmarking Long-Term Memory for Personalized Agents