Morocco

Large Language Models (LLMs) are expected to generate appropriate responses while adhering to predefined prior constraints or knowledge, such as user personas, across various dialogue scenarios. However, real-world interactions frequently involve semantic conflicts between such prior information and actual user-provided inputs. Despite this, prior studies on persona-grounded dialogue—one of the representative tasks in personal preference modeling—have predominantly assumed idealized scenarios where persona and user utterances are fully aligned. To bridge this gap, we introduce and formalize the notion of persona conflict, wherein predefined personas contradict the personal information expressed by the user during interaction. We present a systematic verification framework to examine model behavior under such conflict scenarios. In detail, we propose a taxonomy that categorizes model behaviors into three distinct response types (adhering, sycophantic, and wavering) and develop a measurement schema grounded in this taxonomy. Our study provides a comprehensive analysis of the persona conflict phenomenon, identifying diverse key behavioral factors. Extensive experiments and in-depth analysis provide new insights into designing robust dialogue models capable of managing persona inconsistencies.

EACL 2026

I Know, but I Don&#39;t Know! How Persona Conflict Undermines Instruction Adherence in Large Language Models

behavior style taxonomy

persona conflict problem

verification framework

persona-grounded dialogue

large language models

### Welcome to 2025 Annual Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics

Welcome to the 2025 meeting of the Nations of the Americas Chapter of the Association for Computational Linguistics! I am proud to help organize the first NAACL conference to carry the new name of our organization, one that emphasizes inclusion for all of the Americas. I am also pleased to welcome you to Albuquerque, New Mexico, a state whose unique blend of cultural influences will make for an excellent backdrop for NAACL 2025, especially with this year’s special theme on NLP in a Multicultural World. 
**[Continue reading...](https://drive.google.com/file/d/1jX-qGhqVSZZCIrAnJaz798pflu5Irrdn/view?usp=sharing)**

*- Colin Cherry, Google, NAACL 2025 General Chair* 

[![](https://assets.underline.io/markdown_image/1/image/b087f8a4dc5816d6e1a6514e59c59ac3.png)](https://drive.google.com/file/d/1T96GzPqObXrMln2BMByCSXTSizjTg69P/view?usp=sharing)

You need to log in with the email address you registered with. Access credentials have been sent to your email. 

Please be sure to check your spam and other email folders if you do not see an email confirmation right away.

Please log in to explore this event.

To access NAACL 2025 event page you are required to register. Please follow [**this link**](https://2025.naacl.org/registration/registration/) to register. Access will depend on your registration type.

Please register!

NAACL 2025

Welcome to 2025 Annual Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics

I Know, but I Don't Know! How Persona Conflict Undermines Instruction Adherence in Large Language Models

technical paper

#### *Message from the General Chair, Aline Villavicencio*
I’m delighted and honoured to welcome you to the 19th Conference of the European Chapter of the Association for Computational Linguistics (EACL 2026), taking place in the beautiful city of Rabat, in Morocco, in March 24-29, 2026. EACL is the flagship European conference of the Association and EACL 2026 proudly continues our field’s tradition of excellence in scholarship, innovation, and inclusivity. I am deeply grateful to the many volunteers whose dedication, generosity, and tireless efforts have made this conference possible.
For the first time EACL is being hosted in the African continent. This is an important milestone for our community, and we are grateful to our Moroccan hosts for enabling this historic moment by bringing this edition of EACL to Rabat. We are also delighted that the Second Arabic NLP School is co-located with EACL. We hope attendees enjoy this wonderful opportunity to strengthen ties with the Computational Linguistics communities across the African continent. *[Read full message](https://drive.google.com/file/d/14NlmHvwM6fPJuMmOvVh7K0vtQbEyv3SZ/view?usp=sharing)* 

<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://underline.io/events/522/reception','_blank')">Go to Workshops and Tutorials Program</button></html>

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 EACL 2026". 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 are required to register.
Please complete your registration to continue.

We recommend reading [**the registration information**](https://2026.eacl.org/registration/) first.

**Online Registration Form**: https://acl.swoogo.com/eacl2026

Registration Required

Welcome to the 19th Conference of the European Chapter of the Association for Computational Linguistics (EACL). EACL 2026 will be held in Rabat, Morocco, from March 24–29, 2026. 

Large language models (LLMs) have been widely deployed in Conversational AIs (CAIs), while exposing privacy and security threats. Recent research shows that LLM-based CAIs can be manipulated to extract private information from human users, posing serious security threats. However, the methods proposed in that study rely on a white-box setting that adversaries can directly modify the system prompt. This condition is unlikely to hold in real-world deployments. The limitation raises a critical question: can unprivileged attackers still induce such privacy risks in practical LLM-integrated applications? To address this question, we propose VortexPIA, a novel indirect prompt injection attack that induces privacy extraction in LLM-integrated applications under black-box settings. By injecting token-efficient data containing false memories, VortexPIA misleads LLMs to actively request private information in batches. Unlike prior methods, VortexPIA allows attackers to flexibly define multiple categories of sensitive data. We evaluate VortexPIA on six LLMs, covering both traditional and reasoning LLMs, across four benchmark datasets. The results show that VortexPIA significantly outperforms baselines and achieves state-of-the-art (SOTA) performance. It also demonstrates efficient privacy requests, reduced token consumption, and enhanced robustness against defense mechanisms. We further validate VortexPIA on multiple realistic open-source LLM-integrated applications, demonstrating its practical effectiveness.

VortexPIA: Indirect Prompt Injection Attack against LLMs for Efficient Extraction of User Privacy

Speculative Decoding (SD) has emerged as a widely used paradigm to accelerate the inference of large language models (LLMs) without compromising generation quality. It works by efficiently drafting multiple tokens using a compact model and then verifying them in parallel using the target LLM. Notably, Self-Speculative Decoding proposes skipping certain layers to construct the draft model, which eliminates the need for additional parameters or training. Despite its strengths, we observe in this work that drafting with layer skipping exhibits significant sensitivity to domain shifts, leading to a substantial drop in acceleration performance. To enhance the domain generalizability of this paradigm, we introduce KNN-SSD, an algorithm that leverages K-Nearest Neighbor (KNN) search to match different skipped layers with various domain inputs. We evaluated our algorithm in various models and multiple tasks, observing that its application leads to 1.3x 1.6x speedup in LLM inference.

KNN-SSD: Enabling Dynamic Self-Speculative Decoding via Nearest Neighbor Layer Set Optimization

Large language models (LLMs) are increasingly embedded in Computer Science (CS) classrooms to automate code generation, feedback, and assessment. However, their susceptibility to adversarial or ill-intentioned prompts threatens student learning and academic integrity. To cope with this important issue, we evaluate existing off-the-shelf LLMs in handling unsafe and irrelevant prompts within the domain of CS education. We identify important shortcomings in existing LLM guardrails which motivates us to propose CodeGuard, a comprehensive guardrail framework for educational AI systems. CodeGuard includes (i) a first-of-its-kind taxonomy for classifying prompts; (ii) the CodeGuard dataset, a collection of 8,000 prompts spanning the taxonomy; and (iii) PromptShield, a lightweight sentence-encoder model fine-tuned to detect unsafe prompts in real time. Experiments show that PromptShield achieves 0.93 F1 score, surpassing existing guardrail methods. Additionally, further experimentation reveals that CodeGuard reduces potentially harmful or policy-violating code completions by 30-65% without degrading performance on legitimate educational tasks. The code, datasets, and evaluation scripts are made freely available to the community.

CodeGuard: Improving LLM Guardrails in CS Education

The performance of large language models (LLMs) tends to degrade for extremely low-resource languages, primarily due to the lack of labeled training data. Despite growing interest, the availability of high-quality natural language processing (NLP) datasets for these languages remains limited. This paper addresses such gap by focusing on Ladin, an endangered Romance language, specifically the Val Badia variant. Leveraging a small set of parallel LadinвЂ“Italian sentence pairs, we create synthetic datasets for sentiment analysis and question answering by translating monolingual Italian data. To ensure linguistic quality, we apply rigorous filtering and back-translation procedures in our method. We further demonstrate that incorporating these synthetic datasets into machine translation training leads to substantial improvements over existing ItalianвЂ“Ladin translation baselines. Our contributions include sentiment analysis and question answering datasets for Ladin, establishing foundational resources that support broader NLP research and downstream applications for underrepresented languages.

Towards the First NLP Benchmark for Ladin - an Extremely Low-Resource Language

Activation steering or editing hidden states to control language-model behavior can be framed as a causal mediation problem: inputs induce internal activations, a subset of which act as mediators transmitting targeted behaviors to outputs. We formalize a structural graph over transformer layers and derive front-door—style identification conditions that justify steering through mediating subspaces while preserving non-mediating features, thereby reducing confounding and off-target effects. Within this mediation-first view, we present CAS-BiPO, a sparse mediation steering approach that learns targeted behavioral interventions via regularized training. Empirically, our method achieves 97-100% of dense baseline effectiveness across four behavioral control tasks while using only 10-30% of activation dimensions. Learned masks concentrate 94.3% of steering effects in 26.7% of dimensions, with neurons exhibiting 2.2 x higher activation changes, validating the sparse mediation hypothesis. Our causal framework provides theoretical grounding while CAS-BiPO demonstrates that end-to-end learning of interpretable, reliable interventions is both feasible and advantageous.

Causal Activation Steering via Sparse Mediation

Direct Preference Optimization (DPO) is a powerful approach for aligning large language models (LLMs) with human preferences by formulating preference learning as a supervised classification problem over pairwise human-labeled outputs, thereby enabling stable and efficient training. We show that DPO inherits bias from confounders (e.g., topic, style, user objectives) that shape data generation and carry through to training, hindering recovery of true human preferences. We address this from a causal perspective, proposing Causal Direct Preference Optimization (CDPO), a general framework that incorporates causal inference principles to mitigate the influence of confounders and sharpen the signal of genuine human preferences. Our approach preserves the tractability of direct optimization while enhancing robustness to spurious correlations and annotation biases. Empirical evaluations on benchmark datasets show that CDPO surpasses DPO-based baselines by achieving unbiased fine-tuning through causal reasoning, confirming the effectiveness of confounder-aware preference optimization.

Causal Direct Preference Optimization for Language Model Alignment

Prompts are the interface for eliciting the capabilities of large language models (LLMs). Understanding their structure and components is critical for analyzing LLM behavior and optimizing performance. However, the field lacks a comprehensive framework for systematic prompt analysis and understanding. We introduce PromptPrism, a linguistically-inspired taxonomy that enables prompt analysis across three hierarchical levels: functional structure, semantic component, and syntactic pattern. By applying linguistic concepts to prompt analysis, PromptPrism bridges traditional language understanding and modern LLM research, offering insights that purely empirical approaches might miss. We show the practical utility of PromptPrism by applying it to three applications: (1) a taxonomy-guided prompt refinement approach that automatically improves prompt quality and enhances model performance across a range of tasks; (2) a multi-dimensional dataset profiling method that extracts and aggregates structural, semantic, and syntactic characteristics from prompt datasets, enabling comprehensive analysis of prompt distributions and patterns; (3) a controlled experimental framework for prompt sensitivity analysis by quantifying the impact of semantic reordering and delimiter modifications on LLM performance. Our experimental results validate the effectiveness of our taxonomy across these applications, demonstrating that PromptPrism provides a foundation for refining, profiling, and analyzing prompts.

PromptPrism: A Linguistically-Inspired Taxonomy for Prompts

The strong capabilities of recent Large Language Models (LLMs) have made them highly effective for zero-shot re-ranking task. Attention-based re-ranking methods, which derive relevance scores directly from attention weights, offer an efficient and interpretable alternative to generation-based re-ranking methods. However, they still face two major limitations. First, attention signals are highly concentrated a small subset of tokens within a few documents, making others indistinguishable. Second, attention often overemphasizes phrases lexically similar to the query, yielding biased rankings that irrelevant documents with mere lexical resemblance are regarded as relevant. In this paper, we propose \textbf{ReAttn}, a post-hoc re-weighting strategy for attention-based re-ranking methods. It first compute the cross-document IDF weighting to down-weight attention on query-overlapping tokens that frequently appear across the candidate documents, reducing lexical bias and emphasizing distinctive terms. It then employs entropy-based regularization to mitigate over-concentrated attention, encouraging a more balanced distribution across informative tokens. Both adjustments operate directly on existing attention weights without additional training or supervision. Extensive experiments demonstrate the effectiveness of our method.

ReAttn: Improving Attention-based Re-ranking via Attention Re-weighting

Large Language Models (LLMs) are increasingly deployed in domains such as education, mental health and customer support, where stable and consistent personas are critical for reliability. Yet, existing studies focus on narrative or role-playing tasks and overlook how adversarial conversational history alone can reshape induced personas. Black-box persona manipulation remains unexplored, raising concerns for robustness in realistic interactions. In response, we introduce the task of persona editing, which adversarially steers LLM traits through user-side inputs under a black-box, inference-only setting. To this end, we propose PHISH (Persona Hijacking via Implicit Steering in History), the first framework to expose a new vulnerability in LLM safety that embeds semantically loaded cues into user queries to gradually induce reverse personas. We also define a metric to quantify attack success. Across 3 benchmarks and 8 LLMs, PHISH predictably shifts personas, triggers collateral changes in correlated traits, and exhibits stronger effects in multi-turn settings. In high-risk domains mental health, tutoring, and customer support, PHISH reliably manipulates personas, validated by both human and LLM-as-Judge evaluations. Importantly, PHISH causes only a small reduction in reasoning benchmark performance, leaving overall utility largely intact while still enabling significant persona manipulation. While current guardrails offer partial protection, they remain brittle under sustained attack. Our findings expose new vulnerabilities in personas and highlight the need for context-resilient persona in LLMs. Our codebase and dataset is available at: https://github.com/Jivnesh/PHISH

Persona Jailbreaking in Large Language Models

The paradigm of augmenting Large Language Models (LLMs) with external tools is rapidly converging on the Model-Context Protocol (MCP), a federated architecture of tool "servers'' that mirrors the modern internet. Existing benchmarks, however, remain misaligned with this architecture, typically evaluating agents with APIs in an unstructured, non-hierarchical manner. They often test components like retrieval and reasoning in isolation or sidestep the critical real-world challenge of functional overlap among tools. This leads to an incomplete and overly optimistic assessment of agent capabilities. To address these gaps, we introduceMSC-Bench, a large-scale benchmark for evaluating end-to-end, multi-hop tool use within an explicit, hierarchical MCP ecosystem. A core contribution is our methodology for constructing ground truth by identifying "equal function sets''---groups of functionally equivalent tools. This approach enables the use of objective and reproducible metrics (e.g., F1-score) and mitigates the reliance on fallible LLM-as-a-judge evaluations. MSC-Bench is structured as a five-level curriculum designed to systematically test a full spectrum of agent capabilities, from single-tool execution to complex cross-server orchestration and robustness to out-of-scope requests. Our experiments reveal that rigid hierarchies can hinder performance without co-designed strategies and that even state-of-the-art agents exhibit systemic weaknesses in robustness. MSC-Bench provides a much-needed diagnostic framework to expose these limitations and guide the development of more capable and efficient tool-using systems.

Downloads

Next from EACL 2026

VortexPIA: Indirect Prompt Injection Attack against LLMs for Efficient Extraction of User Privacy

Similar lecture

Dialogue Language Model with Large-Scale Persona Data Engineering