United States

Embedding models are often used for semantic retrieval in high-stakes domains such as law enforcement, where biased outputs can have severe consequences. We systematically measure racial bias in six widely used embedding models by computing similarity scores between crime incident texts that include racial identity tokens and simple law enforcement queries. The analysis reveals that racial descriptors consistently affect cosine similarity scores and retrieval rankings for semantically identical crime incidents. All models exhibit statistically significant bias, with magnitude varying across models. This study provides a comprehensive methodology and metrics to aid the selection of embedding models when deploying NLP-based systems in the law enforcement domain. Organizations can reduce bias at low cost through informed model selection. The methodology establishes reproducible metrics for measuring bias in embedding-based systems.

ACL 2026

Measuring and Mitigating Racial Bias in Embedding Models: A Comparative Study for Law Enforcement Retrieval

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.

poster

While Large Language Models (LLMs) are increasingly integrated into in-vehicle conversational systems, identifying the optimal model remains challenging due to the lack of domain-specific evaluation standards tailored to real-world deployment requirements. In this paper, we propose a novel evaluation framework for in-vehicle assistants, with a particular focus on Korean-language localization. Our empirical analysis reveals notable patterns in model behavior. First, fine-grained Korean honorific control remains unstable in current LLMs, indicating that precise speech-level realization must be explicitly evaluated in localization settings. Second, models exhibit weaker performance in strategic conversational metrics like clarification and proactivity. Our analysis suggests this stems from the inherent subjective complexity of these tasks, where our framework adopts a conservative evaluation stance to prioritize reliability. Together, our findings underscore that automotive AI must move beyond general competence toward precise linguistic tailoring and reliable, safety-oriented interaction management.

LoCar: Localization-Aware Evaluation of In-Vehicle Assistants through Fine-Grained Sociolinguistic Control

Recent advances in large language models (LLMs) have shown promise in formal theorem proving, yet evaluating semantic correctness remains challenging. 
Existing evaluations rely on indirect proxies such as a lexical overlap to human-annotated proof, or expensive manual inspection.
Inspired by the shift from lexical comparison to test-case **execution**-based evaluation in code generation, we propose TE$^2$, a framework that evaluates *semantic correctness* of formal theorems: a generated theorem is considered correct only if all dependent *downstream theorems* compile successfully, analogous to integration testing.
We construct a benchmark from **6** real-world Lean 4 repositories, comprising **2,206 problems** paired with **41** downstream theorems on average, automatically extracted without human effort.
Experiments demonstrate that while state-of-the-art models achieve high compilation success, they perform significantly worse under our semantic metric.
The best model, Claude-Sonnet-4.5, achieves only 38.9% Execution Accuracy in the hard set, given both natural language proof and downstream theorems as context, revealing a critical gap in current theorem generation capabilities.

Benchmarking Testing in Automated Theorem Proving

As Multi-Agent Systems (MAS) scale to hundreds of specialized agents, the routing layer becomes a critical bottleneck. Traditional approaches force a stark trade-off: prompt-based LLM routers deliver high semantic reasoning but incur prohibitive latency (~1500–2000ms) and cost that grows with agent count, while embedding-based routers operate at low latency (25–50ms on a T4 for cosine/centroid-style routing) but fail to capture nuanced functional intent, collapsing semantically similar but functionally distinct agents. We identify representation anisotropy, the geometric collapse of hidden-state vectors into a narrow cone, as a key mechanism underlying this embedding-based routing failure. We propose LatentGate, a non-generative routing architecture that extracts mean-pooled hidden states from a frozen small language model (SLM), applies PCA-whitening (decorrelation + variance normalization) to resolve the anisotropy, and trains a lightweight linear probe for agent classification. Experiments across 5 SLM backbones and 100 enterprise agents show that LatentGate achieves 98.8% in-domain accuracy and 80.0% out-of-distribution accuracy on natural queries, outperforming embedding-based routers by 13–22 absolute points. LatentGate operates at ~28ms on a T4 GPU; the SLM forward pass is independent of agent count, with classification adding an $O(Ck)$ term (agents $C$, whitened dimension $k$) that is negligible at $C = 100$ and small relative to the SLM forward pass. The lightweight linear probe additionally enables sub-10ms warm-start retraining from user feedback, offering a path toward self-healing routing in production. We further benchmark prompt-based routing with GPT-4.1, GPT-4.1-nano, and Gemini 2.5 Flash, demonstrating that these degrade to 70–77% accuracy at 100 agents while incurring 1500–2000ms latency, confirming the need for non-generative alternatives. We categorize this work as Emerging, as it introduces a new routing primitive rather than reporting a completed long-term deployment study.

LatentGate: Low-Latency Semantic Routing via Frozen-Backbone Probing of Small Language Models

Interpreto is an open-source Python library for interpreting HuggingFace language models, from early BERT variants to LLMs. It provides two complementary families of methods: attribution methods and concept-based explanations. The library bridges recent research and practical tooling by exposing explanation workflows through a unified API for both classification and text generation. A key differentiator is its end-to-end concept-based pipeline (from activation extraction to concept learning, interpretation, and scoring), which goes beyond feature-level attributions and is uncommon in existing libraries.

Interpreto: An Explainability Library for Transformers

We present TruthSplit, an interactive system for comparative argument analysis across ideological worldviews. While existing argumentation tools assess arguments against a single standard of correctness, TruthSplit models conditional validity: how the validity of a claim depends on worldview-specific values, assumptions, and conceptual definitions.
Given an input argumentative text, TruthSplit extracts claims and premises, applies three-layer natural language inference (NLI) to assess logical and normative consistency, and conditions large language model (LLM) reasoning on structured worldview profiles that encode core values and decision principles. The system generates worldview-specific interpretations, identifies value conflicts and assumption gaps, and visualizes divergence through interactive analytical interfaces.
Thus, by integrating argument mining, NLI-based consistency modeling, structured knowledge representation, and LLM-conditioned reasoning into a unified framework for multi-perspective argument analysis, TruthSplit enables a systematic comparison of how different worldviews interpret the same argument.

TruthSplit: Revealing Conditional Validity in Arguments Through Multi-Worldview Comparative Reasoning

Financial reporting systems increasingly use large language models (LLMs) to extract and summarize corporate disclosures. However, most assume a single-market setting and do not address structural differences across jurisdictions. Variations in accounting taxonomies, tagging infrastructures (e.g., XBRL vs. PDF), and aggregation conventions make cross-jurisdiction reporting a semantic alignment and verification challenge.
We present FinReporting, an agentic workflow for localized cross-jurisdiction financial reporting. The system builds a unified canonical ontology over Income Statement, Balance Sheet, and Cash Flow, and decomposes reporting into auditable stages including filing acquisition, extraction, canonical mapping, and anomaly logging. Rather than using LLMs as free-form generators, FinReporting deploys them as constrained verifiers under explicit decision rules and evidence grounding.
Evaluated on annual filings from the US, Japan, and China, the system improves consistency and reliability under heterogeneous reporting regimes. We release an interactive demo supporting cross-market inspection and structured export of localized financial statements.

FinReporting: An Agentic Workflow for Localized Reporting of Cross-Jurisdiction Financial Disclosure

Agentic systems are becoming more capable: agents define strategies, take actions, and interact with different environments. This freedom poses serious challenges to overseeing and assessing the agent's behavior. Most current tools are limited, focusing on observability with basic evaluation or creating a static taxonomy of agent errors. To address this gap, we present Agentic CLEAR, an automatic, dynamic, and easy-to-use evaluation framework. It produces textual insights into the agent behavior on three levels of granularity: system, trace, and node. Agentic CLEAR operates above the observability layer, enabling seamless integration and featuring an intuitive UI that makes agent evaluation highly accessible. In our experiments on four benchmarks, seven agentic settings, and tens of thousands of LLM calls, we show that Agentic CLEAR produces high-quality, data-dependent, insightful feedback. Our analysis shows strong alignment with human-annotated errors and the ability to predict task success rate.

Agentic CLEAR: Automating Multi-Level Evaluation of LLM Agents

Clinicians exploring oncology trial repositories often need ad-hoc, multi-constraint queries over biomarkers, endpoints, interventions, and time, yet writing SQL requires schema expertise. We demo FD-NL2SQL, a feedback-driven clinical NL2SQL assistant for SQLite-based oncology databases. Given a natural-language question, a schema-aware LLM decomposes it into predicate-level sub-questions, retrieves semantically similar expert-verified NL2SQL exemplars via sentence embeddings, and synthesizes executable SQL conditioned on the decomposition, retrieved exemplars, and schema, with post-processing validity checks. To improve with use, FD-NL2SQL incorporates two update signals: (i) clinician edits of generated SQL are approved and added to the exemplar bank; and (ii) lightweight logic-based SQL augmentation applies a single atomic mutation (e.g., operator or column change), retaining variants only if they return non-empty results. A second LLM generates the corresponding natural-language question and predicate decomposition for accepted variants, automatically expanding the exemplar bank without additional annotation. The demo interface exposes decomposition, retrieval, synthesis, and execution results to support interactive refinement and continuous improvement.

FD-NL2SQL: Feedback-Driven Clinical NL2SQL that Improves with Use

Diagram question answering (Diagram QA) requires reasoning-level attribution that links each question–answer pair to all visual regions needed to derive the answer, rather than only the region containing the final response. Creating such structured evidence across diagrams, charts, maps, circuits, and infographics is time-consuming, and existing annotation tools tightly couple their interfaces to dataset-specific formats. We present **DIAGRAMS**, a lightweight, schema-driven review framework that decouples interface logic from dataset-specific JSON structures through an internal meta-schema and dataset adapters. Given an image and QA pair with optional candidate regions, the system performs QA-conditioned evidence selection and proposes the regions required for reasoning. When QA pairs or candidate regions are missing, it generates them and supports human verification and refinement. Across six Diagram QA datasets, model-suggested evidence achieves 85.39\% precision and 75.30\% recall against reviewer-final selections (micro-averaged). These results indicate that the review-first framework reduces manual region creation while maintaining high agreement with final reasoning-level attributions. We release a public demo and installable package to support dataset auditing, grounded supervision creation, and grounded evaluation.

DIAGRAMS : A Review Framework for Reasoning-Level Attribution in Diagram QA

High-quality time-aligned annotation is fundamental to speech processing and animal vocalization research, yet precise boundary localization and consistent labeling remain challenging in collaborative settings. We present Praat++, a web-based multimedia annotation system designed for collaborative, video-informed, and AI-assisted timeline labeling of audio and video data. The system tightly synchronizes waveform, spectrogram, pitch, intensity, and time-aligned video playback with fine-grained region-based editing, enabling precise boundary refinement and improved label accuracy within a unified interface. Praat++ further incorporates role-aware workflow management and human-in-the-loop AI-assisted pre-annotation to enhance inter-annotator consistency and reduce labeling time. Through real-world multimodal speech and animal vocalization annotation scenarios, we demonstrate that Praat++ provides an integrated infrastructure for improving annotation quality and efficiency in dataset construction workflows. The demo video (https://www.youtube.com/watch?v=YboCoBRF5lg) and website (https://redgiant.uta.edu/praat) are now publicly available.

Downloads

Next from ACL 2026

LoCar: Localization-Aware Evaluation of In-Vehicle Assistants through Fine-Grained Sociolinguistic Control

Similar lecture

Thesis Proposal: Auditing and Mitigating Demographic Bias in Multi-Stage Retrieval Systems for Criminal Justice Applications