Token Reduction Should Go Beyond Efficiency in Generative Models -- From Vision, Language to Multimodality

• URL: http://arxiv.org/abs/2505.18227v2
• Date: Mon, 28 Jul 2025 01:59:08 GMT
• Title: Token Reduction Should Go Beyond Efficiency in Generative Models -- From Vision, Language to Multimodality
• Authors: Zhenglun Kong, Yize Li, Fanhu Zeng, Lei Xin, Shvat Messica, Xue Lin, Pu Zhao, Manolis Kellis, Hao Tang, Marinka Zitnik,
• Abstract summary: This paper argues that token reduction should transcend its traditional efficiency-oriented role in the era of large generative models.<n>We argue that token reduction can facilitate deeper multimodal integration and alignment, maintain coherence over long inputs, and enhance training stability.<n>We outline promising future directions, including algorithm design, reinforcement learning-guided token reduction, token optimization for in-context learning, and broader ML and scientific domains.
• Score: 29.531450446701175
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: In Transformer architectures, tokens\textemdash discrete units derived from raw data\textemdash are formed by segmenting inputs into fixed-length chunks. Each token is then mapped to an embedding, enabling parallel attention computations while preserving the input's essential information. Due to the quadratic computational complexity of transformer self-attention mechanisms, token reduction has primarily been used as an efficiency strategy. This is especially true in single vision and language domains, where it helps balance computational costs, memory usage, and inference latency. Despite these advances, this paper argues that token reduction should transcend its traditional efficiency-oriented role in the era of large generative models. Instead, we position it as a fundamental principle in generative modeling, critically influencing both model architecture and broader applications. Specifically, we contend that across vision, language, and multimodal systems, token reduction can: (i) facilitate deeper multimodal integration and alignment, (ii) mitigate "overthinking" and hallucinations, (iii) maintain coherence over long inputs, and (iv) enhance training stability, etc. We reframe token reduction as more than an efficiency measure. By doing so, we outline promising future directions, including algorithm design, reinforcement learning-guided token reduction, token optimization for in-context learning, and broader ML and scientific domains. We highlight its potential to drive new model architectures and learning strategies that improve robustness, increase interpretability, and better align with the objectives of generative modeling.

Related papers

• Forgetting: A New Mechanism Towards Better Large Language Model Fine-tuning [53.398270878295754]
  Supervised fine-tuning (SFT) plays a critical role for pretrained large language models (LLMs)<n>We suggest categorizing tokens within each corpus into two parts -- positive and negative tokens -- based on whether they are useful to improve model performance.<n>We conduct experiments on well-established benchmarks, finding that this forgetting mechanism not only improves overall model performance and also facilitate more diverse model responses.
  arXiv  Detail & Related papers  (2025-08-06T11:22:23Z)
• Efficient Machine Unlearning via Influence Approximation [75.31015485113993]
  Influence-based unlearning has emerged as a prominent approach to estimate the impact of individual training samples on model parameters without retraining.<n>This paper establishes a theoretical link between memorizing (incremental learning) and forgetting (unlearning)<n>We introduce the Influence Approximation Unlearning algorithm for efficient machine unlearning from the incremental perspective.
  arXiv  Detail & Related papers  (2025-07-31T05:34:27Z)
• Efficient Attention Mechanisms for Large Language Models: A Survey [18.86171225316892]
  Transformer-based architectures have become the prevailing computation backbone of large language models.<n>Recent research has introduced two principal categories of efficient attention mechanisms.<n>Sparse attention techniques, in contrast, limit attention to selected subsets of tokens based on fixed patterns, block-wise routing, or clustering strategies.
  arXiv  Detail & Related papers  (2025-07-25T18:08:10Z)
• Modality Agnostic Efficient Long Range Encoder [14.705955027331674]
  We address the challenge of long-context processing on a single device using generic implementations.<n>To overcome these limitations, we propose MAELRE, a unified and efficient transformer architecture.<n>We demonstrate that MAELRE achieves superior accuracy while reducing computational cost compared to existing long-context models.
  arXiv  Detail & Related papers  (2025-07-25T16:19:47Z)
• Token Communication in the Era of Large Models: An Information Bottleneck-Based Approach [55.861432910722186]
  UniToCom is a unified token communication paradigm that treats tokens as the fundamental units for both processing and wireless transmission.<n>We propose a generative information bottleneck (GenIB) principle, which facilitates the learning of tokens that preserve essential information.<n>We employ a causal Transformer-based multimodal large language model (MLLM) at the receiver to unify the processing of both discrete and continuous tokens.
  arXiv  Detail & Related papers  (2025-07-02T14:03:01Z)
• Curse of High Dimensionality Issue in Transformer for Long-context Modeling [31.257769500741006]
  We propose textitDynamic Group Attention (DGA) to reduce redundancy by aggregating less important tokens during attention computation.<n>Our results show that our DGA significantly reduces computational costs while maintaining competitive performance.
  arXiv  Detail & Related papers  (2025-05-28T08:34:46Z)
• Enhancing Transformers Through Conditioned Embedded Tokens [28.80560770188464]
  We develop a theoretical framework that establishes a direct relationship between the conditioning of the attention block and that of the embedded tokenized data.<n>We introduce conditioned tokens, a method that systematically modifies the embedded tokens to improve the conditioning of the attention mechanism.<n>Our analysis demonstrates that this approach significantly mitigates ill-conditioning, leading to more stable and efficient training.
  arXiv  Detail & Related papers  (2025-05-19T07:21:53Z)
• Quantifying Memory Utilization with Effective State-Size [73.52115209375343]
  We develop a measure of textitmemory utilization'<n>This metric is tailored to the fundamental class of systems with textitinput-invariant and textitinput-varying linear operators
  arXiv  Detail & Related papers  (2025-04-28T08:12:30Z)
• Over-Tokenized Transformer: Vocabulary is Generally Worth Scaling [10.985444895887207]
  We introduce Over-Tokenized Transformers, a framework that decouples input and output vocabularies to improve language modeling performance.<n>We uncover a log-linear relationship between input vocabulary size and training loss, demonstrating that larger input vocabularies consistently enhance model performance.<n>Our findings highlight the importance of tokenization in scaling laws and provide practical insight for tokenizer design.
  arXiv  Detail & Related papers  (2025-01-28T14:15:42Z)
• Autonomous Structural Memory Manipulation for Large Language Models Using Hierarchical Embedding Augmentation [0.0]
  This study introduces hierarchical embedding augmentation as a means to redefine the representation of tokens through multi-level semantic structures.<n>Results reveal substantial improvements in computational efficiency, with marked reductions in processing overhead for longer input sequences.<n>The ability to dynamically adjust token representations and memory configurations contributed to the model's robustness under varied and unpredictable input conditions.
  arXiv  Detail & Related papers  (2025-01-23T22:20:36Z)
• 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.<n>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)
• big.LITTLE Vision Transformer for Efficient Visual Recognition [34.015778625984055]
  big.LITTLE Vision Transformer is an innovative architecture aimed at achieving efficient visual recognition. System is composed of two distinct blocks: the big performance block and the LITTLE efficiency block. When processing an image, our system determines the importance of each token and allocates them accordingly.
  arXiv  Detail & Related papers  (2024-10-14T08:21:00Z)
• Semantic Equitable Clustering: A Simple and Effective Strategy for Clustering Vision Tokens [57.37893387775829]
  We introduce a fast and balanced clustering method, named textbfSemantic textbfEquitable textbfClustering (SEC) SEC clusters tokens based on their global semantic relevance in an efficient, straightforward manner. We propose a versatile vision backbone, SECViT, to serve as a vision language connector.
  arXiv  Detail & Related papers  (2024-05-22T04:49:00Z)
• Interpreting and Improving Attention From the Perspective of Large Kernel Convolution [51.06461246235176]
  We introduce Large Kernel Convolutional Attention (LKCA), a novel formulation that reinterprets attention operations as a single large- Kernel convolution.<n>LKCA achieves competitive performance across various visual tasks, particularly in data-constrained settings.
  arXiv  Detail & Related papers  (2024-01-11T08:40:35Z)
• GroupBERT: Enhanced Transformer Architecture with Efficient Grouped Structures [57.46093180685175]
  We demonstrate a set of modifications to the structure of a Transformer layer, producing a more efficient architecture. We add a convolutional module to complement the self-attention module, decoupling the learning of local and global interactions. We apply the resulting architecture to language representation learning and demonstrate its superior performance compared to BERT models of different scales.
  arXiv  Detail & Related papers  (2021-06-10T15:41:53Z)
• Improve Variational Autoencoder for Text Generationwith Discrete Latent Bottleneck [52.08901549360262]
  Variational autoencoders (VAEs) are essential tools in end-to-end representation learning. VAEs tend to ignore latent variables with a strong auto-regressive decoder. We propose a principled approach to enforce an implicit latent feature matching in a more compact latent space.
  arXiv  Detail & Related papers  (2020-04-22T14:41:37Z)

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.