China

Aligning large language models to handle instructions with extremely long contexts has yet to be fully investigated. Previous studies have attempted to scale up the available data volume by synthesizing long instruction-following samples, as constructing such a dataset tends to be challenging for annotators. However, a lack of a well-defined strategy for ensuring data quality may introduce low-quality samples and restrict the model’s performance. Thus, we propose GATEAU, a novel framework to address the unique challenge of long context alignment by identifying the influential samples enriched with long-range dependency relations. Specifically, GATEAU measures the long-range dependencies from two essential aspects: the difficulty of generating target responses due to the long-range dependencies, and the difficulty of understanding long inputs due to such dependencies. Comprehensive experiments indicate that GATEAU effectively identifies influential samples and the model trained on these selected samples exhibits better instruction-following and long-context understanding capabilities.

EMNLP 2025

GATEAU: Selecting Influential Samples for Long Context Alignment

long-context lm

llms

alignment

### Welcome to The 63rd Annual Meeting of the Association for Computational Linguistics (ACL 2025)

Message from the General Chair: 
*It is my great pleasure and honor to welcome you to the 63rd Annual Meeting of the Association for Computational Linguistics (ACL 2025), held in beautiful Vienna, Austria, from July 27 to August 1, 2025. ACL2025continues our field’s 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.* 
[Read more](https://drive.google.com/file/d/1GI_hvOpjswAuYdUTromfeDiPpCcqidwg/view?usp=sharing)

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 ACL 2025". 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 need to log in with the **email address you registered with**. 

ACL 2025

Welcome to The 63rd Annual Meeting of the Association for Computational Linguistics (ACL 2025)

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.

Multimodal large language models (MLLMs) demonstrate impressive capabilities by integrating visual and textual information. However, the incorporation of visual modalities also introduces new and complex safety risks, rendering even the most advanced models vulnerable to sophisticated jailbreak attacks. This paper first analyzes the impact of inserting safety reasoning prompts on various aspects of the model. We find that this external method can help the model resist jailbreak attacks to some extent, but the model still fails to distinguish specific semantic scenarios, resulting in a significantly increased refusal rate for benign queries. Inspired by this, we propose a novel training framework, \textbf{SURE} (Safety Understanding and Reasoning Enhancement for Multimodal Large Language Models), designed to help models internalize chain-of-thought-based safety decision-making capabilities. Extensive experiments demonstrate that SURE significantly improves model safety while effectively avoiding over-defense, achieving a good balance between safety and generality. Finally, we create a large-scale multimodal safety reasoning dataset, MLLM-SCoT-Plus, to facilitate research on safety alignment in multimodal models. Our data and code are publicly available at \url{https://anonymous.4open.science/r/SURE}.

SURE: Safety Understanding and Reasoning Enhancement for Multimodal Large Language Models

Experts in machine learning leverage domain knowledge to navigate decisions in model selection, hyperparameter optimisation, and resource allocation. This is particularly critical for fine-tuning language models (LMs), where repeated trials incur substantial computational overhead and environmental impact. However, no existing automated framework simultaneously tackles the entire model selection and HPO task for resource-efficient LM fine-tuning. We introduce XAutoLM, a meta-learning–augmented AutoML framework that reuses past experiences to optimize discriminative and generative LM fine-tuning pipelines efficiently. XAutoLM learns from stored successes and failures by extracting task- and system-level meta-features to bias its sampling toward fruitful configurations and away from costly dead ends. On four text classification and two question-answering benchmarks, XAutoLM surpasses zero-shot optimiser’s peak F1 on five of six tasks, cuts mean evaluation time by up to 4.5×, reduces error ratios by up to sevenfold, and uncovers up to 50% more pipelines above the zero-shot Pareto front. In contrast, simpler memory-based baselines suffer negative transfer. We release XAutoLM and our experience store to catalyze resource-efficient, Green AI fine-tuning in the NLP community.

XAutoLM: Efficient Fine-Tuning of Language Models via Meta-Learning and AutoML

With rapid advancement and increasing accessibility of LLMs, fine-tuning aligned models has become a critical step for adapting them to real-world applications, which makes the safety of this fine-tuning process more important than ever. However, recent studies have highlighted a critical challenge: even when fine-tuning with seemingly benign downstream datasets, the safety of aligned LLMs can be compromised, making them more susceptible to malicious instructions. In this paper, we show that fine-tuning datasets often contain samples with safety-degrading features that are not easily identifiable on the surface. These samples can significantly degrade the safety alignment of LLMs during fine-tuning. To address this issue, we propose LARF, a \textbf{L}ayer-\textbf{A}ware \textbf{R}epresentation \textbf{F}iltering method. This method identifies safety-sensitive layers within the LLM and leverages their representations to detect which data samples in the post-training dataset contain safety-degrading features. Experimental results demonstrate that LARF can effectively identify benign data with safety-degrading features. After removing such data, the safety alignment degradation caused by fine-tuning is mitigated.

Layer-Aware Representation Filtering: Purifying Finetuning Data to Preserve LLM Safety Alignment

Small language models (SLMs) are gaining attention for their lower computational and memory needs while maintaining strong performance. However, efficiently deploying SLMs on resource-constrained devices remains a significant challenge. Post-training quantization (PTQ) is a widely used compression technique that reduces memory usage and inference computation, yet existing methods face challenges in inefficient bit-width allocation and insufficient fine-grained quantization adjustments, leading to suboptimal performance, particularly at lower bit-widths. To address these challenges, we propose multi-level weight quantization (MLWQ), which facilitates the efficient deployment of SLMs. Our method enables more effective bit-width allocation by jointly considering inter-layer loss and intra-layer salience. Furthermore, we propose a fine-grained partitioning of intra-layer salience to support the tweaking of quantization parameters within each group. Experimental results show that MLWQ achieves superior performance compared to state-of-the-art methods, providing an effective approach for the efficient deployment of SLMs while maintaining model accuracy.

MLWQ: Efficient Small Language Model Deployment via Multi-Level Weight Quantization

Recent advances show that the world knowledge in the Instruction Fine-Tuning (IFT) dataset, which is incompatible with LLMs' internal knowledge, can greatly hurt the IFT performance. However, the effective integration and balancing of the internal knowledge of LLMs, acquired during pre-training, with existing IFT datasets remains a largely underexplored area of research. To address this gap, this work introduces NILE, a novel framework to optimize the effectiveness of IFT by adjusting IFT datasets through carefully aligning the world and internal knowledge. NILE employs a three-stage pipeline to effectively quantify and adjust consistency with the internal knowledge of target LLMs. Our analysis provides compelling evidence that balancing such consistency with pre-trained internal knowledge is pivotal for unleashing LLM potential, and confirms that NILE can systematically contribute to these substantial performance improvements. Experimental results demonstrate that NILE-aligned IFT datasets sharply boost LLM performance across multiple LLM ability evaluation datasets, achieving up to 66.6% gain on Arena-Hard and 68.5% on Alpaca-Eval V2.

NILE: Internal Consistency Alignment in Large Language Models

Large Language Models (LLMs) are increasingly used to generate synthetic textual data for training smaller specialized models. However, a comparison of various generation strategies for low-resource language settings is lacking. While various prompting strategies have been proposed—such as demonstrations, label-based summaries, and self-revision—their comparative effectiveness remains unclear, especially for low-resource languages. In this paper, we systematically evaluate the performance of these generation strategies and their combinations across 11 typologically diverse languages, including several extremely low-resource ones. Using three NLP tasks and four open-source LLMs, we assess downstream model performance on generated versus gold-standard data. Our results show that strategic combinations of generation methods — particularly target-language demonstrations with LLM-based revisions — yield strong performance, narrowing the gap with real data to as little as 5% in some settings. We also find that smart prompting techniques can reduce the advantage of larger LLMs, highlighting efficient generation strategies for synthetic data generation in low-resource scenarios with smaller models.

A Rigorous Evaluation of LLM Data Generation Strategies for Low-Resource Languages

Privacy-sensitive users require deploying large language models (LLMs) within their own infrastructure (\textit{on-premises}) to safeguard private data and enable customization. However, vulnerabilities in local environments can lead to unauthorized access and potential model theft. To address this, prior research on small models has explored securing only the output layer within hardware-secured devices to balance model confidentiality and customization. Yet this approach fails to protect LLMs effectively. In this paper, we discover that (1) query-based distillation attacks targeting the secured top layer can produce a functionally equivalent replica of the victim model; (2) securing the same number of layers, bottom layers before a transition layer provide stronger protection against distillation attacks than top layers, with comparable effects on customization performance; and (3) the number of secured layers creates a trade-off between protection and customization flexibility. Based on these insights, we propose SOLID, a novel deployment framework that secures a few bottom layers in a secure environment and introduces an efficient metric to optimize the trade-off by determining the ideal number of hidden layers. Extensive experiments on five models (1.3B to 70B parameters) demonstrate that SOLID outperforms baselines, achieving a better balance between protection and downstream customization.

A Middle Path for On-Premises LLM Deployment: Preserving Privacy Without Sacrificing Model Confidentiality

Large Language Models (LLMs) with extended context lengths face significant computational challenges during the pre-filling phase, primarily due to the quadratic complexity of self-attention. Existing methods typically employ dynamic pattern matching and block-sparse low-level implementations. However, their reliance on local information for pattern identification fails to capture global contexts, and the coarse granularity of blocks leads to persistent internal sparsity, resulting in suboptimal accuracy and efficiency. To address these limitations, we propose \textbf{AnchorAttention}, a difference-aware, dynamic sparse attention mechanism that efficiently identifies critical attention regions at a finer stripe granularity while adapting to global contextual information, achieving superior speed and accuracy. AnchorAttention comprises three key components: (1) \textbf{Pattern-based Anchor Computation}, leveraging the commonalities present across all inputs to rapidly compute a set of near-maximum scores as anchor; (2) \textbf{Difference-aware Stripe Sparsity Identification}, performing difference-aware comparisons with anchor to quickly obtain discrete coordinates of significant regions in a stripe-like sparsity pattern; (3) \textbf{Fine-grained Sparse Computation}, replacing the traditional contiguous loading strategy with a discrete key-value loading approach to maximize sparsity rates while preserving hardware computational potential. Additionally, we integrate the identification strategy into a single operator to maximize parallelization potential. With its finer-grained sparsity strategy, \textbf{AnchorAttention} achieves higher sparsity rates at the same recall level, significantly reducing computation time. Compared to previous state-of-the-art methods, at a text length of 128k, it achieves a speedup of 1.44times while maintaining higher recall rates.

AnchorAttention: Difference-Aware Sparse Attention with Stripe Granularity

Fine-tuning pre-trained large language models (LLMs) on downstream tasks has achieved significant success across various domains. However, as model sizes grow, traditional first-order fine-tuning algorithms incur substantial memory overhead due to the need for activation storage for back-propagation (BP). The BP-free Memory-Efficient Zeroth-Order Optimization (MeZO) method estimates gradients through finite differences, avoiding the storage of activation values, and has been demonstrated as a viable approach for fine-tuning large language models. This work proposes the Multiple-query Memory Efficient Zeroth-Order (MUZO) method, which is based on variance-reduced multiple queries to obtain the average of gradient estimates. When combined with Adam optimizer, MUZO-Adam demonstrates superior performance in fine-tuning various LLMs. Furthermore, we provide theoretical guarantees for the convergence of the MUZO-Adam optimizer. Extensive experiments empirically demonstrate that MUZO-Adam converges better than MeZO-SGD and achieves near first-order optimizer performance on downstream classification, multiple-choice, and generation tasks.

MUZO: Leveraging Multiple Queries and Momentum for Zeroth-Order Fine-Tuning of Large Language Models

Vision Language Models (VLMs) face significant computational challenges with long-form videos due to attention mechanisms' quadratic complexity. We propose Language-Guided Temporal Token Pruning (LGTTP), which extracts temporal cues from queries to adaptively assign pruning rates across video frames, preserving contextual continuity while reducing computational overhead. Unlike existing methods that apply uniform token pruning or disruptive keyframe selection, LGTTP maintains higher token density in temporally relevant segments. Our model-agnostic approach integrates with TimeChat and LLaVA-Video architectures, demonstrating substantial efficiency gains across diverse benchmarks. Experiments show LGTTP reduces computation by 65% while preserving 97-99% of original performance. On highlight detection, LGTTP achieves +9.5% HIT@1 on QVHighlights compared to baseline methods, and for temporal grounding, it maintains 99.6% of original R@1 on Charades-STA. LGTTP shows particular strength for queries with explicit temporal markers while remaining effective for general video understanding tasks.

Downloads

Next from EMNLP 2025

SURE: Safety Understanding and Reasoning Enhancement for Multimodal Large Language Models

Similar lecture

LADM: Long-context Training Data Selection with Attention-based Dependency Measurement for LLMs