SAISA: Towards Multimodal Large Language Models with Both Training and Inference Efficiency

• URL: http://arxiv.org/abs/2502.02458v1
• Date: Tue, 04 Feb 2025 16:28:53 GMT
• Title: SAISA: Towards Multimodal Large Language Models with Both Training and Inference Efficiency
• Authors: Qianhao Yuan, Yanjiang Liu, Yaojie Lu, Hongyu Lin, Ben He, Xianpei Han, Le Sun,
• Abstract summary: We introduce SAISA, a novel architecture that enhance both training and inference efficiency.<n>Using the same configuration as LLaVA-1.5, SAISA reduces inference FLOPs by 66% and training budget by 26%, while achieving superior performance in terms of accuracy.
• Score: 47.03718208259308
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Multimodal Large Language Models (MLLMs) mainly fall into two architectures, each involving a trade-off between training and inference efficiency: embedding space alignment (e.g., LLaVA-1.5) is inefficient during inference, while cross-attention space alignment (e.g., Flamingo) is inefficient in training. In this paper, we compare these two architectures and identify the key factors for building efficient MLLMs. A primary difference between them lies in how attention is applied to visual tokens, particularly in their interactions with each other. To investigate whether attention among visual tokens is necessary, we propose a new self-attention mechanism, NAAViT (\textbf{N}o \textbf{A}ttention \textbf{A}mong \textbf{Vi}sual \textbf{T}okens), which eliminates this type of attention. Our pilot experiment on LLaVA-1.5 shows that attention among visual tokens is highly redundant. Based on these insights, we introduce SAISA (\textbf{S}elf-\textbf{A}ttention \textbf{I}nput \textbf{S}pace \textbf{A}lignment), a novel architecture that enhance both training and inference efficiency. SAISA directly aligns visual features with the input spaces of NAAViT self-attention blocks, reducing computational overhead in both self-attention blocks and feed-forward networks (FFNs). Using the same configuration as LLaVA-1.5, SAISA reduces inference FLOPs by 66\% and training budget by 26\%, while achieving superior performance in terms of accuracy. Comprehensive ablation studies further validate the effectiveness of SAISA across various LLMs and visual encoders. The code and model will be publicly available at https://github.com/icip-cas/SAISA.

Related papers

• SemanticVLA: Semantic-Aligned Sparsification and Enhancement for Efficient Robotic Manipulation [65.6201974979119]
  We propose SemanticVLA, a novel VLA framework that performs Semantic-Hierarchical Sparsification and Enhancement for Efficient Robotic Manipulation.<n>SemanticVLA surpasses OpenVLA on LIBERO benchmark by 21.1% in success rate, while reducing training cost and inference latency by 3.0-fold and 2.7-fold.
  arXiv  Detail & Related papers  (2025-11-13T17:24:37Z)
• $\mathcal{V}isi\mathcal{P}runer$: Decoding Discontinuous Cross-Modal Dynamics for Efficient Multimodal LLMs [26.779915891040236]
  We propose emphVisiPruner, a training-free pruning framework that reduces up to 99% of vision-related attention computations and 53.9% of FLOPs on LLaVA-v1.5 7B.<n>Our insights further provide actionable guidelines for training efficient MLLMs by aligning model architecture with its intrinsic layer-wise processing dynamics.
  arXiv  Detail & Related papers  (2025-10-20T06:40:17Z)
• Don't Just Chase "Highlighted Tokens" in MLLMs: Revisiting Visual Holistic Context Retention [50.97683288777336]
  Multimodal Large Language Models (MLLMs) suffer from considerable computational overhead due to their reliance on massive visual tokens.<n>Recent studies have explored token pruning to alleviate this problem, which typically uses text-vision cross-attention.<n>We propose HoloV, a plug-and-play visual token pruning framework for efficient inference.
  arXiv  Detail & Related papers  (2025-10-03T11:33:40Z)
• Generic Token Compression in Multimodal Large Language Models from an Explainability Perspective [6.258220461022373]
  Existing Multimodal Large Language Models (MLLMs) process a large number of visual tokens, leading to significant computational costs and inefficiency.<n>We show that token compression is feasible at the input stage of LLM with negligible performance loss.<n>We propose to learn a mapping from the attention map of the first LLM layer to the explanation results, thereby avoiding the need for a full inference pass.
  arXiv  Detail & Related papers  (2025-06-01T17:44:16Z)
• DeepInsert: Early Layer Bypass for Efficient and Performant Multimodal Understanding [26.39397960987363]
  We propose a simple modification to pretrained transformer models.<n>Instead of concatenation with the language prompt at the start, we insert multimodal tokens directly into the middle.<n>Our results indicate that our method reduces computational costs during both training and inference.
  arXiv  Detail & Related papers  (2025-04-27T18:56:26Z)
• Beyond Token Compression: A Training-Free Reduction Framework for Efficient Visual Processing in MLLMs [38.34856927170692]
  Multimodal Large Language Models (MLLMs) are typically based on decoder-only or cross-attention architectures.<n>They require significantly higher computational resources due to extensive self-attention and FFN operations on visual tokens.<n>We present a novel analysis framework to investigate the necessity of these costly operations in decoder-only MLLMs.
  arXiv  Detail & Related papers  (2025-01-31T11:09:16Z)
• ST$^3$: Accelerating Multimodal Large Language Model by Spatial-Temporal Visual Token Trimming [14.937905258757635]
  $textbfST3$ is a framework designed to accelerate MLLM inference without retraining.<n>$textbfST3$ can be seamlessly integrated into existing pre-trained MLLMs.
  arXiv  Detail & Related papers  (2024-12-28T10:17:29Z)
• Core Context Aware Attention for Long Context Language Modeling [50.774702091154204]
  We propose a plug-and-play Core Context Aware (CCA) Attention for efficient long-range context modeling. Our CCA-Attention significantly outperforms state-of-the-art models in terms of computational efficiency and long-context modeling ability.
  arXiv  Detail & Related papers  (2024-12-17T01:54:08Z)
• SepLLM: Accelerate Large Language Models by Compressing One Segment into One Separator [65.62084602011596]
  Large Language Models (LLMs) have exhibited exceptional performance across a spectrum of natural language processing tasks.<n>We have identified a key pattern: certain seemingly meaningless special tokens (i.e., separators) contribute disproportionately to attention scores compared to semantically meaningful tokens.<n>We introduce SepLLM, a plug-and-play framework that accelerates inference by compressing these segments and eliminating redundant tokens.
  arXiv  Detail & Related papers  (2024-12-16T18:58:57Z)
• Compressing KV Cache for Long-Context LLM Inference with Inter-Layer Attention Similarity [24.118503938098307]
  Existing methods, including selective token retention and window-based attention, improve efficiency but risk discarding important tokens needed for future text generation.<n>We propose an approach that enhances LLM efficiency without token loss by reducing the memory and computational load of less important tokens, rather than discarding them.
  arXiv  Detail & Related papers  (2024-12-03T08:29:27Z)
• Skipping Computations in Multimodal LLMs [63.29737699997859]
  This study investigates redundancy in Multimodal Large Language Models (MLLMs) during inference. We propose different methods to skip computations, such as skipping entire blocks, FFN or self-attention layers. Our findings validate that significant amount of computations can be avoided at inference time.
  arXiv  Detail & Related papers  (2024-10-12T09:21:45Z)
• PAR: Prompt-Aware Token Reduction Method for Efficient Large Multimodal Models [32.33892531885448]
  Multimodal large language models (MLLMs) demonstrate strong performance across visual tasks.<n>But their efficiency is hindered by significant computational and memory demands from processing long contexts in multimodal inputs.<n>We introduce PAR (Prompt-Aware Token Reduction), a novel and plug-and-play approach that reduces visual tokens efficiently without compromising model performance.
  arXiv  Detail & Related papers  (2024-10-09T07:13:22Z)
• EchoAtt: Attend, Copy, then Adjust for More Efficient Large Language Models [29.57891007810509]
  Large Language Models (LLMs) have demonstrated outstanding performance across a variety of natural language processing tasks. We introduce EchoAtt, a novel framework aimed at optimizing transformer-based models by analyzing and leveraging the similarity of attention patterns across layers. Our best results with TinyLLaMA-1.1B demonstrate that EchoAtt improves inference speed by 15%, training speed by 25%, and reduces the number of parameters by approximately 4%, all while improving zero-shot performance.
  arXiv  Detail & Related papers  (2024-09-22T21:08:37Z)
• EE-MLLM: A Data-Efficient and Compute-Efficient Multimodal Large Language Model [15.449472477182061]
  Current approaches for vision and language interaction fall into two categories: self-attention-based and cross-attention-based methods. We modify the original self-attention mechanism in MLLM to a composite attention mechanism. EE-MLLM significantly outperforms Flamingo with limited training data, and reduces the prefilling time to 79 ms on an H800 GPU. We present a training-free variant named EE-MLLM-F, which reduces the computation cost of self-attention-based method without additional training.
  arXiv  Detail & Related papers  (2024-08-21T17:36:37Z)
• AsyCo: An Asymmetric Dual-task Co-training Model for Partial-label Learning [53.97072488455662]
  Self-training models achieve state-of-the-art performance but suffer from error accumulation problem caused by mistakenly disambiguated instances. We propose an asymmetric dual-task co-training model called AsyCo, which forces its two networks, i.e., a disambiguation network and an auxiliary network, to learn from different views explicitly. Experiments on both uniform and instance-dependent partially labeled datasets demonstrate the effectiveness of AsyCo.
  arXiv  Detail & Related papers  (2024-07-21T02:08:51Z)
• Not All Attention is Needed: Parameter and Computation Efficient Transfer Learning for Multi-modal Large Language Models [73.48675708831328]
  We propose a novel parameter and computation efficient tuning method for Multi-modal Large Language Models (MLLMs) The Efficient Attention Skipping (EAS) method evaluates the attention redundancy and skips the less important MHAs to speed up inference. The experiments show that EAS not only retains high performance and parameter efficiency, but also greatly speeds up inference speed.
  arXiv  Detail & Related papers  (2024-03-22T14:20:34Z)
• Efficient Streaming Language Models with Attention Sinks [72.20260088848987]
  StreamingLLM is an efficient framework that enables Large Language Models to generalize to infinite sequence lengths without any fine-tuning. We show that StreamingLLM can enable Llama-2, MPT, Falcon, and Pythia to perform stable and efficient language modeling with up to 4 million tokens and more.
  arXiv  Detail & Related papers  (2023-09-29T17:59:56Z)
• MA2CL:Masked Attentive Contrastive Learning for Multi-Agent Reinforcement Learning [128.19212716007794]
  We propose an effective framework called textbfMulti-textbfAgent textbfMasked textbfAttentive textbfContrastive textbfLearning (MA2CL) MA2CL encourages learning representation to be both temporal and agent-level predictive by reconstructing the masked agent observation in latent space. Our method significantly improves the performance and sample efficiency of different MARL algorithms and outperforms other methods in various vision-based and state-based scenarios.
  arXiv  Detail & Related papers  (2023-06-03T05:32:19Z)
• Towards Joint Intent Detection and Slot Filling via Higher-order Attention [47.78365472691051]
  Intent detection (ID) and Slot filling (SF) are two major tasks in spoken language understanding (SLU) We propose a Bilinear attention block, which exploits both the contextual and channel-wise bilinear attention distributions. We show that our approach yields improvements compared with the state-of-the-art approach.
  arXiv  Detail & Related papers  (2021-09-18T09:50:23Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.