United States

Long-context Multimodal Large Language Models (MLLMs) that incorporate long text-image and text-video modalities, demand substantial computational resources as their multimodal Key-Value (KV) cache grows with increasing input lengths, challenging memory and time efficiency. For multimodal scenarios, the cross-modal interactions inevitablely increase complexity, and prior methods for KV cache compression, in both text-only and multimodal LLMs, have neglected attention density variations across layers, often adopting uniform or progressive reduction strategis for layer-wise cache allocation. This results in precision loss and suboptimal performance. We propose \textbf{MEDA}, a novel approach specifically designed for the complexities of multimodal settings, dynamically allocating KV cache sizes based on attention entropy to better adapt to multimodal interactions.Through a dynamic multimodal KV cache allocation strategy, MEDA compresses the KV cache, adaptively retains sufficient multimodal information at each layer. Meanwhile, to mitigate the degradation of contextual information due to cache compression, we also integrate KV pairs merging techniques to maintain coherence. MEDA achieves up to \textbf{72\%} KV cache memory reduction and \textbf{2.82} faster decoding speeds in some cases, while maintaining or enhancing performance on various multimodal tasks in a long context, including multi-image and long video scenarios.

NAACL 2025

MEDA: Dynamic KV Cache Allocation for Efficient Multimodal Long-Context Inference

multimodal long-context

kv cache

poster

### 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!

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

Query-focused tabular summarization is an emerging task in table-to-text generation that synthesizes a summary response from tabular data based on user queries. Traditional transformer-based approaches face challenges due to token limitations and the complexity of reasoning over large tables. To address these challenges, we introduce DETQUS (Decomposition-Enhanced Transformers for QUery-focused Summarization), a system designed to improve summarization accuracy by leveraging tabular decomposition alongside a fine-tuned encoder-decoder model. DETQUS employs a large language model to selectively reduce table size, retaining only query-relevant columns while preserving essential information. This strategy enables more efficient processing of large tables and enhances summary quality. Our approach, equipped with table-based QA model Omnitab, achieves a ROUGE-L score of 0.4437, outperforming the previous state-ofthe- art REFACTOR model (ROUGE-L: 0.422). These results highlight DETQUS as a scalable and effective solution for query-focused tabular summarization, offering a structured alternative to more complex architectures.

DETQUS: Decomposition-Enhanced Transformers for QUery-focused Summarization

Recent advancements in Large Language Models (LLMs) have set themselves apart with their exceptional performance in complex language modelling tasks. However, these models are also known for their significant computational and storage requirements, primarily due to the quadratic computation complexity of softmax attention. To mitigate this issue, linear attention has been designed to reduce the quadratic space-time complexity that is inherent in standard transformers. In this work, we embarked on a comprehensive exploration of three key components that substantially impact the performance of the Gated Linear Attention module: feature maps, normalization, and the gating mechanism. We developed a feature mapping function to address some crucial issues that previous suggestions overlooked. Then we offered further rationale for the integration of normalization layers to stabilize the training process. Moreover, we explored the saturation phenomenon of the gating mechanism and augmented it with a refining module. We conducted extensive experiments and showed our architecture outperforms previous Gated Linear Attention mechanisms in extensive tasks including training from scratch and post-linearization with continual pre-training.

ReGLA: Refining Gated Linear Attention

Recent progress in video-text retrieval has been driven largely by advancements in model architectures and training strategies. However, the representation learning capabilities of video-text retrieval models remain constrained by low-quality and limited training data annotations. To address this issue, we present a novel Vi\textsc{\textbf{d}}eo-Text \textsc{\textbf{R}}etrieval Paradigm with R\textsc{\textbf{e}}levance-based \textsc{\textbf{A}}ug\textsc{\textbf{m}}entation, namely \textsc{dReAm}, which enhances video and text data using large foundation models to learn more generalized features. Specifically, we first adopt a simple augmentation method, which generates self-similar data by randomly duplicating or dropping subwords and frames. In addition, inspired by the recent advancement in visual and language generative models, we propose a more robust augmentation method through textual paraphrasing and video stylization using large language models (LLMs) and visual generative models (VGMs). To further enrich video and text information, we propose a relevance-based augmentation method, where LLMs and VGMs generate and integrate new relevant information into the original data. Leveraging this enriched data, extensive experiments on several video-text retrieval benchmarks demonstrate the superiority of \textsc{dReAm} over existing methods. Code will be available upon acceptance.

DREAM: Improving Video-Text Retrieval Through Relevance-Based Augmentation Using Large Foundation Models

Retrieval-Augmented Generation (RAG) models excel in knowledge-intensive tasks, especially under few-shot learning constraints. We introduce CoRAG, a framework extending RAG to collaborative settings, where clients jointly train a shared model using a collaborative passage store. To evaluate CoRAG, we introduce CRAB, a benchmark for collaborative homogeneous open-domain question answering. Our experiments demonstrate that CoRAG consistently outperforms both parametric collaborative learning methods and locally trained RAG models in low-resource scenarios. Further analysis reveals the critical importance of relevant passages within the shared store, the surprising benefits of incorporating irrelevant passages, and the potential for hard negatives to negatively impact performance. This introduces a novel consideration in collaborative RAG: the trade-off between leveraging a collectively enriched knowledge base and the potential risk of incorporating detrimental passages from other clients. Our findings underscore the viability of CoRAG, while also highlighting key design challenges and promising avenues for future research.

CoRAG: Collaborative Retrieval-Augmented Generation

There is a significant gap between patient needs and available mental health support today. In this paper, we aim to thoroughly examine the potential of using Large Language Models (LLMs) to assist professional psychotherapy. To this end, we propose a new benchmark, CBT-Bench, for the systematic evaluation of cognitive behavioral therapy (CBT) assistance. We include three levels of tasks in CBT-Bench: **I: Basic CBT knowledge acquisition**, with the task of multiple-choice questions; **II: Cognitive model understanding**, with the tasks of cognitive distortion classification, primary core belief classification, and fine-grained core belief classification; **III: Therapeutic response generation**, with the task of generating responses to patient speech in CBT therapy sessions. These tasks encompass key aspects of CBT that could potentially be enhanced through AI assistance, while also outlining a hierarchy of capability requirements, ranging from basic knowledge recitation to engaging in real therapeutic conversations. We evaluated representative LLMs on our benchmark. Experimental results indicate that while LLMs perform well in reciting CBT knowledge, they fall short in complex real-world scenarios requiring deep analysis of patients' cognitive structures and generating effective responses, suggesting potential future work.

CBT-Bench: Evaluating Large Language Models on Assisting Cognitive Behavior Therapy

Humans are accustomed to reading and writing in a forward manner, and this natural bias extends to text understanding in auto-regressive large language models (LLMs). This paper investigates whether LLMs, like humans, struggle with reverse modeling, specifically with reversed text inputs. We found that publicly available pre-trained LLMs cannot understand such inputs. However, LLMs trained from scratch with both forward and reverse texts can understand them equally well during inference across multiple languages. Our case study shows that different-content texts result in different losses if input (to LLMs) in different directions---some get lower losses for forward while some for reverse. This leads us to a simple and nice solution for data selection based on the loss differences between forward and reverse directions. Using our selected data in continued pretraining can boost LLMs' performance by a large margin across different language understanding benchmarks.

Reverse Modeling in Large Language Models

Despite significant advancements, the practical deployment of Large Language Models (LLMs) is often hampered by their immense sizes, highlighting the need for effective compression techniques. Singular Value Decomposition (SVD) emerges as a promising method for compressing LLMs. However, existing SVD-based compression approaches suffer from substantial truncation losses, leading to severe performance degradation in compressed models. In this work, we introduce \sysname, a novel SVD-based LLM compression method that optimizes singular value truncation in SVD compression with two key strategies. First, \sysname employs dynamic compression ratio allocation to effectively balance the extremely large truncation loss across different layers. Second, it implements loss-optimized weight truncation to ensure that the truncated singular values result in a lower and more stable truncation loss in practice. We evaluate \sysname on ten datasets and five models on various scales and demonstrated that \sysname outperforms current state-of-the-art methods. The source code is available at \url{https://github.com/AIoT-MLSys-Lab/SVD-LLM}.

SVD-LLM V2: Optimizing Singular Value Truncation for Large Language Model Compression

Missing modality issues are common in real-world applications, arising from factors such as equipment failures and privacy concerns. When fine-tuning pre-trained models on downstream datasets with missing modalities, performance can degrade significantly. Current methods often aggregate various missing cases to train recovery modules or align multimodal features, resulting in suboptimal performance, high computational costs, and the risk of catastrophic forgetting in continual environments where data arrives sequentially. In this paper, we formulate the dynamic missing modality problem as a continual learning task and introduce the continual multimodal missing modality task. To address this challenge efficiently, we introduce three types of prompts: modality-specific, task-aware, and task-specific prompts. These prompts enable the model to learn intra-modality, inter-modality, intra-task, and inter-task features. Furthermore, we propose a contrastive task interaction strategy to explicitly learn prompts correlating different modalities. We conduct extensive experiments on three public datasets, where our method consistently outperforms state-of-the-art approaches.

Efficient Prompting for Continual Adaptation to Missing Modalities

Emergent abilities of large language models (LLMs) have significantly advanced their application in autonomous vehicle (AV) research. Safe integration of LLMs into vehicles, however, necessitates their thorough understanding of dynamic traffic environments. Towards this end, this study introduces a framework leveraging LLMs' built-in extrapolation capabilities for vehicle trajectory prediction, thereby evaluating their comprehension of the evolution of traffic agents' behaviors and interactions over time. The framework employs a traffic encoder to extract spatial-level scene features from agents' observed trajectories to facilitate efficient scene representation. To focus on LLM's innate capabilities, scene features are then converted into LLM-compatible tokens through a reprogramming adapter and finally decoded into predicted trajectories with a linear decoder. Experimental results quantitatively demonstrate the framework's efficacy in enabling off-the-shelf, frozen LLMs to achieve competitive trajectory prediction performance, with qualitative analyses revealing their enhanced understanding of complex, multi-agent traffic scenarios.

Harnessing and Evaluating the Intrinsic Extrapolation Ability of Large Language Models for Vehicle Trajectory Prediction

Recent advanced frameworks in topic models have significantly enhanced the performance compared to conventional probabilistic approaches. Such models, mostly constructed from neural network architecture together with other advanced techniques such as contextual embedding, optimal transport distance and pre-trained language model, etc. have effectively improved the topic quality and document topic distribution. Despite the improvements, these methods lack considerations of effective optimization for complex objective functions that contain log-likelihood and additional regularization terms. In this study, we propose to apply an efficient optimization method to improve the generalization and performance of topic models. Our approach explicitly considers the sharpness of the loss landscape during optimization, which forces the optimizer to choose directions in the parameter space that lead to flatter minima, in which the models are typically more stable and robust to small perturbations in the data. Additionally, we propose an effective strategy to select the flatness region for parameter optimization by leveraging the optimal transport distance between doc-topic distributions and doc-cluster proportions, which can effectively enhance document representation. Experimental results on popular benchmark datasets demonstrate that our method effectively improves the performance of baseline topic models.

MEDA: Dynamic KV Cache Allocation for Efficient Multimodal Long-Context Inference

Downloads

Next from NAACL 2025

DETQUS: Decomposition-Enhanced Transformers for QUery-focused Summarization

Stay up to date with the latest Underline news!

PRESENTATIONS

CONFERENCES

COMPANY

RESOURCES