EACL 2026 Papers

As the scale of Large Language Models (LLMs) continues to grow rapidly, the cost of training and inference has significantly increased, limiting their application in resource-constrained scenarios. To address this challenge, model pruning has been widely used to reduce computational complexity. Among various pruning strategies, block-wise pruning has gained popularity due to its ability to accelerate computation by removing entire blocks of parameters. However, existing methods often rely on hard labels from calibration datasets and neglect the cumulative effects of pruning on subsequent blocks. To address this, we propose two complementary techniques: the Logit Disruption Score (LDS), a novel block importance criterion that measures the impact of pruning by comparing the cosine similarity between the logits of the original and pruned models, focusing on the most informative logit dimensions to better preserve the model’s core capabilities; and Activation Statistics Correction (ASC), an affine transformation mechanism that aligns the mean and variance of activations in the pruned model with those of the original model, effectively mitigating the distribution shift caused by block removal and improving the information flow in subsequent blocks. Experiments across multiple datasets show that our approach reduces reliance on calibration data and improves generalization, achieving competitive results with existing methods.

Large language models (LLMs) have shown strong performance in zero-shot summarization, but often struggle to model document structure and identify salient information in long texts. In this work, we introduce StrucSum, a training-free prompting framework that enhances LLM reasoning through sentence-level graph structures. StrucSum injects structural signals into prompts via three targeted strategies: Neighbor-Aware Prompting (NAP) for local context, Centrality-Aware Prompting (CAP) for importance estimation, and Centrality-Guided Masking (CGM) for efficient input reduction. Experiments on ArXiv, PubMed, and Multi-News demonstrate that StrucSum consistently improves both summary quality and factual consistency over unsupervised baselines and vanilla prompting. In particular, on ArXiv, it increases FactCC and SummaC by 19.2% and 8.0% points, demonstrating stronger alignment between summaries and source content. The ablation study shows that the combination of multiple strategies does not yield clear performance gains; therefore, structure-aware prompting with graph-based information represents a promising and underexplored direction for the advancement of zero-shot extractive summarization with LLMs.

Large language models have shown strong reasoning capabilities through chain-structured methods such as Chain-of-Thought. Recent studies optimize thought structures by generating parallel or tree-like structures, switching long and short reasoning modes, or aligning reasoning steps with task performance. However, these approaches mainly rely on previously generated logical directions of the chains, which ignore the unexplored regions of the solution space. Such a phenomenon is denoted as blind spots, which limit the diversity and effectiveness of the reasoning process. To this end, we propose the “Thought Space Explorer” (TSE), a framework for navigating and expanding thought structures to overcome blind spots in LLM reasoning. Our TSE first identifies key nodes with high impact, then generates new nodes by integrating information from multiple chains. Finally, it extends new branches through connection strategies. We conduct a series of experiments on math and QA benchmarks. Compared to existing baseline methods, TSE improves the accuracy of both the final answer and intermediate reasoning steps, while maintaining a better effectiveness-efficiency trade-off for practical deployment.

Achieving pronunciation proficiency in a second language (L2) remains a challenge, despite the development of Computer-Assisted Pronunciation Training (CAPT) systems. Traditional CAPT systems often provide unintuitive feedback that lacks actionable guidance, limiting its effectiveness. Recent advancements in audio-language models (ALMs) offer the potential to enhance these systems by providing more user-friendly feedback. In this work, we investigate ALMs for chat-based pronunciation training by introducing L2-Arctic-plus, an English dataset with detailed error explanations and actionable suggestions for improvement. We benchmark cascaded ASR+LLMs and existing ALMs on this dataset, specifically in detecting mispronunciation and generating actionable feedback. To improve the performance, we further propose to instruction-tune ALMs on L2-Arctic-plus. Experimental results demonstrate that our instruction-tuned models significantly outperform existing baselines on mispronunciation detection and suggestion generation in terms of both objective and human evaluation, highlighting the value of the proposed dataset.

Multi-hop reasoning over long contexts remains challenging, as it requires integrating relevant contexts scattered across distant sources while resisting semantic drift and noise from distracting content. While retrieval-augmented generation (RAG) has emerged as the prevailing solution, most RAG approaches encode and store context in monolithic memory representations, resulting in noisy retrieval and brittle reasoning. To overcome these limitations, we introduce TAG (Tailoring Memory Granularity), a framework that prestructures memory into diverse granularities and employs a reward-guided navigator to adaptively compose hybrid memory tailored to each query. The navigator is trained with a multi-objective Bradley–Terry loss that learns the relative utility of different memory types, enabling dynamic routing across granularities. This design allows RAG systems to balance fine-grained detail with high-level abstraction, yielding more reliable reasoning. Extensive experiments on long-context multi-hop question answering (QA) benchmarks show that TAG achieves state-of-the-art performance. With only 0.033% additional parameters, it remains lightweight, highlighting its practicality as a scalable and effective solution for enhancing language model agents in complex, real-world scenarios.

Illicit drug use among teenagers and young adults (TYAs) remains a pressing public health concern, with rising prevalence and long-term impacts on health and well-being. To detect illicit drug use among TYAs, researchers analyze large-scale surveys such as the Youth Risk Behavior Survey (YRBS) and the National Survey on Drug Use and Health (NSDUH), which preserve rich demographic, psychological, and environmental factors related to substance use. However, existing modeling methods treat survey variables independently, overlooking latent and interconnected structures among them. To address this limitation, we propose LAMI (LAtent relation Mining with bi-modal Interpretability), a novel joint graph-language modeling framework for detecting illicit drug use and interpreting behavioral risk factors among TYAs. LAMI represents individual responses as relational graphs, learns latent connections through a specialized graph structure learning layer, and integrates a large language model to generate natural language explanations grounded in both graph structures and survey semantics. Experiments on the YRBS and NSDUH datasets show that LAMI outperforms competitive baselines in predictive accuracy. Interpretability analyses further demonstrate that LAMI reveals meaningful behavioral substructures and psychosocial pathways, such as family dynamics, peer influence, and school-related distress, that align with established risk factors for substance use. Our codebase is available here.

Dynamic topic models aim to reveal how themes emerge, evolve, and dissolve in time-stamped corpora, but existing approaches still face three major challenges: (i) encoders capture bag-of-words statistics but fail to align with the rich semantic priors of large pre-trained language models, (ii) temporal linkages are often modeled as rigid one-to-one chains, limiting the ability to track non-linear evolution such as topic splits or merges, and (iii) interpretability remains shallow, relying on noisy top-word lists that obscure thematic clarity. We propose L-DNTM (LLM-Augmented for Dynamic Neural Topic Model), a variational framework designed to capture more faithful temporal trajectories. Our model integrates three key components: multi-objective distillation to inject PLM-derived semantic knowledge into the encoder, entropy-regularized optimal transport to align entire topic constellations across time for smooth yet flexible evolution, and LLM-guided refinement to sharpen topic–word distributions for improved interpretability. Extensive experiments on diverse corpora show that L-DNTM yields more coherent, temporally consistent, and interpretable topic dynamics, and further enhances downstream classification and clustering tasks.

The Text-to-SQL task translates natural language questions into SQL queries, enabling intuitive database interaction for non-experts. While recent methods leveraging Large Language Models (LLMs) achieve strong performance, their reliance on proprietary models raises concerns about deployment feasibility and data privacy. In this work, we introduce LitE-SQL, a Lightweight and Efficient framework with two components: (i) a Schema Retriever that performs efficient schema linking using a vector database of pre-computed schema embeddings, optimized with a hard-negative supervised contrastive objective to distinguish semantically similar but functionally irrelevant columns, and (ii) a SQL Generator fine-tuned in two stages—supervised fine-tuning followed by execution-guided reinforcement—enabling execution-guided self-correction without multi-candidate sampling, which is commonly required by prior LLM-based approaches.On BIRD, LitE-SQL achieves 72.10% execution accuracy, and on Spider 1.0 it reaches 88.45%, demonstrating comparable or superior performance to LLM-based methods despite using 2× to 30× fewer parameters. Our findings demonstrate that high-quality Text-to-SQL generation is feasible with lightweight models, offering a practical solution for privacy-sensitive and resource-constrained settings.

Large Reasoning Models (LRMs) are powerful, but they still suffer from inefficient and off-target reasoning. Currently, training-free methods are limited to either rigid heuristics or descriptive, non-actionable analyses. In this paper, we introduce ThinkPilot, a training-free framework that automatically optimizes LRMs reasoning. It uses an evolutionary process to generate think-prefixes, namely instructions that evolve driven by a taxonomy of reasoning behaviors to guide models toward superior performance. Extensive experiments demonstrate ThinkPilot’s broad effectiveness: it significantly improves the accuracy-length trade-off for efficient reasoning, drastically improves safety (e.g., cutting the StrongREJECT score of DeepSeek-R1-Distill-Qwen-32B from 27.0% to 0.7%), and enhances instruction following. It also synergizes with existing training-based methods. Specially, our analysis reveals that think-prefixes can reliably control LRMs’ reasoning behaviors, and that different tasks have strong preferences for specific behavioral distributions. By automatically identifying and eliciting these behaviors, ThinkPilot provides a generalizable framework for aligning LRMs reasoning with task demands.

Preference-based alignment like Reinforcement Learning from Human Feedback (RLHF) learns from pairwise preferences, yet the labels are often noisy and inconsistent. Existing uncertainty-aware approaches weight preferences, but ignore a more fundamental factor: the reliability of the answers being compared. To address the problem, we propose Conformal Feedback Alignment (CFA), a framework that grounds preference weighting in the statistical guarantees of Conformal Prediction (CP). CFA quantifies answer-level reliability by constructing conformal prediction sets with controllable coverage and aggregates these reliabilities into principled weights for both DPO- and PPO-style training. Experiments across different datasets show that CFA improves alignment robustness and data efficiency, highlighting that modeling answer-side uncertainty complements preference-level weighting and yields more robust, data-efficient alignment.

In the tabular domain, which is the predominant data format in real-world applications, anomalies are extremely rare or difficult to collect, as their identification often requires domain expertise. Consequently, evaluating tabular anomaly detection models is challenging, since anomalies may be absent even in evaluation sets. To tackle this challenge, prior works have generated synthetic anomaly generation rely on statistical patterns, they often overlook domain semantics and struggle to reflect the complex, domain-specific nature of real-world anomalies. We propose AutoAnoEval, a novel evaluation framework for tabular AD that constructs pseudo-evaluation sets with semantically grounded synthetic anomalies. Our approach leverages an iterative interaction between a Large Language Model (LLM) and a decision tree (DT): the LLM generates realistic anomaly conditions based on contextual semantics, while the DT provides structural guidance by capturing feature interactions inherent in the tabular data. This iterative loop ensures the generation of diverse anomaly conditions, ranging from easily detectable outliers to subtle cases near the decision boundary. Extensive experiments on 20 tabular AD benchmarks demonstrate that AutoAnoEval achieves superior model selection performance, with high ranking alignment and minimal performance gaps compared to evaluations on anomalies encountered in practical applications.

Different large language models (LLMs) exhibit diverse strengths and weaknesses, and LLM ensemble serves as a promising approach to integrate their complementary capabilities. Despite substantial progress in improving ensemble quality, limited attention has been paid to the robustness of ensembles against potential erroneous signals, which often arise from heterogeneous tokenization schemes and varying model expertise. Our analysis shows that ensemble failures typically arise from both the token level and the model level: the former reflects severe disagreement in token predictions, while the latter involves low confidence and pronounced disparities among models. In light of this, we propose CoRE, a plug-and-play technique that harnesses model consistency for robust LLM ensemble, which can be seamlessly integrated with diverse ensemble methods. Token-level consistency captures fine-grained disagreements by applying a low-pass filter to downweight uncertain tokens with high inconsistency, often due to token misalignment, thereby improving robustness at a granular level. Model-level consistency models global agreement by promoting model outputs with high self-confidence and minimal divergence from others, enhancing robustness at a coarser level. Extensive experiments across diverse benchmarks, model combinations, and ensemble strategies demonstrate that CoRE consistently improves ensemble performance and robustness. Our code is available at https://github.com/zhichenz98/CoRE-EACL26.

Knowledge Distillation (KD) has emerged as a crucial technique for compressing Large Language Models (LLMs). Although existing cross-tokenizer KD methods have made notable progress, their effectiveness remains constrained by suboptimal alignment across sequence and vocabulary levels. To address these limitations, we introduce Dual-Space Weighting and Time-Warped Alignment (DWA-KD), a novel cross-tokenizer distillation framework that enhances token-wise distillation through dual-space entropy-based weighting and achieves precise sequence-level alignment by leveraging both lexical and semantic information. At the token level, DWA-KD maps teacher representations into the student space and vice versa, performing dual-space KD via Kullback–Leibler divergence (KL). The process is modulated by dual-space entropy-based weights that up-weight tokens where the student is uncertain and the teacher is confident, thereby focusing learning on informative tokens rather than treating all positions equally. At the sequence level, DWA-KD applies Soft Dynamic Time Warping (Soft-DTW) to both the embedding and final hidden-state layers, enabling robust alignment of lexical and contextual semantics between teacher and student sequences. Extensive experiments across diverse NLP benchmarks demonstrate that DWA-KD consistently outperforms state-of-the-art KD baselines, while ablation studies confirm the complementary contributions of entropy-based token weighting and embedding and final hidden state layer Soft-DTW alignment.

Diffusion language models (DLMs) have recently emerged as a compelling alternative to autoregressive generation, offering parallel generation and improved global coherence. During inference, DLMs generate text by iteratively denoising masked sequences in parallel; however, determining which positions to unmask and which tokens to commit forms a large combinatorial search problem. Existing inference methods approximate this search using heuristics, which often yield suboptimal decoding paths; other approaches instead rely on additional training to guide token selection. To introduce a principled search mechanism for DLMs inference, we introduce MEDAL, an inference-time scaling framework that integrates Monte Carlo Tree SEarch initialization for Diffusion LAnguage Model inference. We employ Monte Carlo Tree Search at the initialization stage to explore promising unmasking trajectories, providing a robust starting point for subsequent refinement. This design enables efficient inference-time scaling, allowing generation quality to improve as the search budget increases, without additional training. Across multiple benchmarks, MEDAL achieves up to 22.0% improvement over existing inference strategies, establishing a new paradigm for search-based inference in DLMs.

Despite remarkable progress in multilingual machine translation (MT), the majority of African—especially East African—languages remain significantly underrepresented both in benchmark datasets and state-of-the-art (SOTA) MT models. This persistent exclusion from mainstream technologies not only limits equitable access, but constrains the development of tools that accurately reflect the region’s linguistic and cultural diversity. Recent advances in open-source large language models have demonstrated strong multilingual MT capabilities through data-efficient adaptation strategies. However, little work has explored their potential for low-resource African languages. We introduce AfriMMT-EA, the first highly multilingual benchmark and MT dataset for East African languages. Our datasets comprise 54 local languages across five East African countries. We used these data to fine-tune two multilingual versions of Gemma-3. We compare models’ performance on these languages with larger off-the-shelf baselines. We release our data and models, in the interest of advancing MT for these low-resource languages and their communities.

Reasoning language models have set state-of-the-art (SOTA) records on many challenging benchmarks, enabled by multi-step reasoning induced by reinforcement learning. However, reasoning models are prone to generating confident, plausible responses that are incorrect (hallucinations). Knowing when and how much to trust these models is critical for safe deployment in real-world applications. To this end, we explore uncertainty quantification (UQ) of reasoning models in this work. We ask three fundamental questions: First, are reasoning models well-calibrated? Second, does deeper reasoning improve model calibration? Finally, inspired by humans’ innate ability to double-check their thought processes to verify the validity of their answers and their confidence, we ask: can reasoning models improve their calibration by explicitly reasoning about their chain-of-thought traces? We introduce introspective uncertainty quantification (IUQ) to explore this direction. In extensive evaluations on SOTA reasoning models across a broad range of benchmarks focused on knowledge-intensive tasks, we find that reasoning models: (i) are typically overconfident, (ii) become even more overconfident with deeper reasoning, and (iii) can become better calibrated through introspection (e.g., o3-Mini and DeepSeek R1) but not uniformly (e.g., Claude 3.7 Sonnet becomes more poorly calibrated). We conclude with important research directions to design necessary UQ benchmarks and improve the calibration of reasoning models.

Dashboards are powerful visualization tools for data-driven decision-making, integrating multiple interactive views that allow users to explore, filter, and navigate data. Unlike static charts, dashboards support rich interactivity, which is essential for uncovering insights in real-world analytical workflows. However, existing question-answering benchmarks for data visualizations largely overlook this interactivity, focusing instead on static charts. This limitation severely constrains their ability to evaluate the capabilities of modern multimodal agents designed for GUI-based reasoning. To address this gap, we introduce DashboardQA, the first benchmark explicitly designed to assess how vision-language GUI agents comprehend and interact with real-world dashboards. The benchmark includes 292 tasks on 112 interactive dashboards, encompassing 405 question answer pairs overall. These questions span five categories: multiple-choice, factoid, hypothetical, multi-dashboard, and conversational. By assessing a variety of leading closed- and open-source GUI agents, our analysis reveals their key limitations, particularly in grounding dashboard elements, planning interaction trajectories, and performing reasoning. Our findings indicate that interactive dashboard reasoning is a challenging task overall for all the VLMs evaluated. Even the top-performing agents struggle; for instance, the best agent based on Gemini-Pro-2.5 achieves only 38.69% accuracy, while the OpenAI CUA agent reaches just 22.69%, demonstrating the benchmark’s significant difficulty. We release DashboardQA at ..

Adapting Large Language Models (LLMs) to specialized domains requires high-quality instruction tuning datasets, which are expensive to create through human annotation. Existing data synthesis methods focus on general-purpose tasks and fail to capture domain-specific terminology and reasoning patterns. To address this, we introduce DS^2-Instruct, a zero-shot framework that generates domain-specific instruction datasets without human supervision. Our approach first generates task-informed keywords to ensure comprehensive domain coverage. It then creates diverse instructions by pairing these keywords with different cognitive levels from Bloom’s Taxonomy. Finally, it uses self-consistency validation to ensure data quality. We apply this framework to generate datasets across seven challenging domains, such as mathematics, finance, and logical reasoning. Comprehensive evaluation demonstrates that models fine-tuned on our generated data achieve substantial improvements over existing data generation methods.

Temporal Knowledge Graph (TKG) reasoning seeks to predict future missing facts from historical evidence. While diffusion models (DM) have recently gained attention for their ability to capture complex predictive distributions, two gaps remain: (i) the generative path is conditioned only on positive evidence, overlooking informative negative context, and (ii) training objectives are dominated by cross-entropy ranking, which improves candidate ordering but provides little supervision over the calibration of the denoised embedding. To bridge this gap, we introduce Negative-Aware Diffusion model for TKG Extrapolation (NADEx). Specifically, NADEx encodes subject-centric histories of entities, relations and temporal intervals into sequential embeddings. NADEx perturbs the query object in the forward process and reconstructs it in reverse with a Transformer denoiser conditioned on the temporal-relational context. We further derive a cosine-alignment regularizer derived from batch-wise negative prototypes, which tightens the decision boundary against implausible candidates. Comprehensive experiments on four public TKG benchmarks demonstrate that NADEx delivers state-of-the-art performance.