Token Reduction

Benchmarks archive 2025-07-28

No benchmark for this task in the archive.

Libraries

Not in the archive: the export carries no per-task library table, so there is nothing to show at snapshot 2025-07-28.

Datasets archive 2025-07-28

No dataset record in the archive lists this task.

Subtasks archive 2025-07-28

No subtask under this task in the archive's task tree.

Parent tasks archive 2025-07-28

Most implemented papers archive 2025-07-28

30 shown of 37 papers with code (78 tagged with this task in all), ordered by repositories listed in the archive, not by stars (the archive holds no stars, so PwC's “Social” and “Latest” sorts cannot be reproduced). Papers without a page here are shown as plain text.

• Learning Compact Vision Tokens for Efficient Large Multimodal Models
  8 Jun 2025 1 repository listed
  Specifically, we propose a Spatial Token Fusion (STF) method to learn compact vision tokens for short vision token sequence, where spatial-adjacent tokens are fused into one.
• SiLVR: A Simple Language-based Video Reasoning Framework
  30 May 2025 1 repository listed
  In the second stage, language descriptions are fed into a powerful reasoning LLM to solve complex video-language understanding tasks.
• One Trajectory, One Token: Grounded Video Tokenization via Panoptic Sub-object Trajectory
  29 May 2025 1 repository listed
  We introduce grounded video tokenization, a paradigm that organizes tokens based on panoptic sub-object trajectories rather than fixed patches.
• FlowCut: Rethinking Redundancy via Information Flow for Efficient Vision-Language Models
  26 May 2025 1 repository listed Syntology ran 1 of 1 samples · 0 unverified
  However, as the interaction between tokens and layers is complicated, this raises a basic question: Is such a simple single-layer criterion sufficient to identify redundancy?
• Token Reduction Should Go Beyond Efficiency in Generative Models -- From Vision, Language to Multimodality
  23 May 2025 1 repository listed
  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.
• Not All Tokens Are What You Need In Thinking
  23 May 2025 1 repository listed Syntology ran 0 of 2 samples · 2 unverified
  Modern reasoning models, such as OpenAI's o1 and DeepSeek-R1, exhibit impressive problem-solving capabilities but suffer from critical inefficiencies: high inference latency, excessive computational resource…
• CrossLMM: Decoupling Long Video Sequences from LMMs via Dual Cross-Attention Mechanisms
  22 May 2025 1 repository listed
  In this paper, we introduce CrossLMM, decoupling long video sequences from LMMs via a dual cross-attention mechanism, which substantially reduces visual token quantity with minimal performance degradation.
• Streamline Without Sacrifice -- Squeeze out Computation Redundancy in LMM
  21 May 2025 1 repository listed Syntology ran 5 of 9 samples · 4 unverified · 4 pointer-only (licence)
  Unlike token reduction methods that focus on token-level redundancy, we identify and study the computation-level redundancy on vision tokens to ensure no information loss.
• PACT: Pruning and Clustering-Based Token Reduction for Faster Visual Language Models
  11 Apr 2025 1 repository listed Syntology ran 1 of 2 samples · 1 unverified
  Visual Language Models require substantial computational resources for inference due to the additional input tokens needed to represent visual information.
• Window Token Concatenation for Efficient Visual Large Language Models
  5 Apr 2025 1 repository listed
  To further enhance the performance on fine-grained visual understanding tasks, we introduce WiCo+, which decomposes the visual tokens in later layers of the LLM.
• Faster Parameter-Efficient Tuning with Token Redundancy Reduction
  26 Mar 2025 1 repository listed
  Parameter-efficient tuning (PET) aims to transfer pre-trained foundation models to downstream tasks by learning a small number of parameters.
• When Large Vision-Language Model Meets Large Remote Sensing Imagery: Coarse-to-Fine Text-Guided Token Pruning
  10 Mar 2025 1 repository listed
  To preserve image details while reducing computational complexity, we propose a text-guided token pruning method with Dynamic Image Pyramid (DIP) integration.
• FOLDER: Accelerating Multi-modal Large Language Models with Enhanced Performance
  5 Jan 2025 1 repository listed
  Recently, Multi-modal Large Language Models (MLLMs) have shown remarkable effectiveness for multi-modal tasks due to their abilities to generate and understand cross-modal data.
• Cached Adaptive Token Merging: Dynamic Token Reduction and Redundant Computation Elimination in Diffusion Model
  1 Jan 2025 1 repository listed Syntology ran 5 of 5 samples · 0 unverified · 5 pointer-only (licence)
  Diffusion models have emerged as a promising approach for generating high-quality, high-dimensional images.
• Rethinking Token Reduction with Parameter-Efficient Fine-Tuning in ViT for Pixel-Level Tasks
  1 Jan 2025 1 repository listed
  Parameter-efficient fine-tuning (PEFT) adapts pre-trained models to new tasks by updating only a small subset of parameters, achieving efficiency but still facing significant inference costs driven by input token length.
• Cross-Layer Cache Aggregation for Token Reduction in Ultra-Fine-Grained Image Recognition
  31 Dec 2024 1 repository listed
  Ultra-fine-grained image recognition (UFGIR) is a challenging task that involves classifying images within a macro-category.
• FrameFusion: Combining Similarity and Importance for Video Token Reduction on Large Visual Language Models
  30 Dec 2024 1 repository listed Syntology ran 2 of 13 samples · 11 unverified
  Leveraging the unique properties of similarity over importance, we introduce FrameFusion, a novel approach that combines similarity-based merging with importance-based pruning for better token reduction in LVLMs.
• Faster Vision Mamba is Rebuilt in Minutes via Merged Token Re-training
  17 Dec 2024 1 repository listed
  Re-training the token-reduced model enhances the performance of Mamba, by effectively rebuilding the key knowledge.
• Enhancing Multimodal Large Language Models Complex Reason via Similarity Computation
  13 Dec 2024 1 repository listed
  Multimodal large language models have experienced rapid growth, and numerous different models have emerged.
• Learning to Merge Tokens via Decoupled Embedding for Efficient Vision Transformers
  13 Dec 2024 1 repository listed Syntology ran 4 of 34 samples · 30 unverified · 4 pointer-only (licence)
  Our method introduces a lightweight embedding module decoupled from the ViT forward pass to extract dedicated features for token merging, thereby addressing the restriction from using intermediate features.
• Token Cropr: Faster ViTs for Quite a Few Tasks
  1 Dec 2024 1 repository listed
  The adoption of Vision Transformers (ViTs) in resource-constrained applications necessitates improvements in inference throughput.
• Inference Optimal VLMs Need Fewer Visual Tokens and More Parameters
  5 Nov 2024 1 repository listed Syntology ran 1 of 2 samples · 1 unverified · 2 pointer-only (licence)
  Our results reveal a surprising trend: for visual reasoning tasks, the inference-optimal behavior in VLMs is achieved by using the largest LLM that fits within the inference budget while minimizing visual token count -…
• LongVU: Spatiotemporal Adaptive Compression for Long Video-Language Understanding
  22 Oct 2024 1 repository listed Syntology ran 6 of 11 samples · 5 unverified
  Given a light-weight LLM, our LongVU also scales effectively into a smaller size with state-of-the-art video understanding performance.
• Rethinking Token Reduction for State Space Models
  16 Oct 2024 1 repository listed Syntology ran 2 of 2 samples · 0 unverified · 2 pointer-only (licence)
  While token reduction techniques offer a straightforward post-training strategy, we find that applying existing methods directly to SSMs leads to substantial performance drops.
• FastAdaSP: Multitask-Adapted Efficient Inference for Large Speech Language Model
  3 Oct 2024 1 repository listed Syntology ran 2 of 3 samples · 1 unverified
  In this study, we aim to explore Multitask Speech Language Model (SpeechLM) efficient inference via token reduction.
• Discovering the Gems in Early Layers: Accelerating Long-Context LLMs with 1000x Input Token Reduction
  25 Sep 2024 1 repository listed Syntology ran 13 of 13 samples · 0 unverified
  Our research introduces a novel approach for the long context bottleneck to accelerate LLM inference and reduce GPU memory consumption.
• Less is More: A Simple yet Effective Token Reduction Method for Efficient Multi-modal LLMs
  17 Sep 2024 1 repository listed Syntology ran 8 of 8 samples · 0 unverified
  The rapid advancement of Multimodal Large Language Models (MLLMs) has led to remarkable performances across various domains.
• Bridging Local Details and Global Context in Text-Attributed Graphs
  18 Jun 2024 1 repository listed Syntology ran 1 of 1 samples · 0 unverified · 1 pointer-only (licence)
  Representation learning on text-attributed graphs (TAGs) is vital for real-world applications, as they combine semantic textual and contextual structural information.
• ALGM: Adaptive Local-then-Global Token Merging for Efficient Semantic Segmentation with Plain Vision Transformers
  14 Jun 2024 1 repository listed Syntology ran 6 of 12 samples · 6 unverified · 12 pointer-only (licence)
  This work presents Adaptive Local-then-Global Merging (ALGM), a token reduction method for semantic segmentation networks that use plain Vision Transformers.
• Hierarchical Context Merging: Better Long Context Understanding for Pre-trained LLMs
  16 Apr 2024 1 repository listed Syntology ran 12 of 15 samples · 3 unverified · 15 pointer-only (licence)
  Large language models (LLMs) have shown remarkable performance in various natural language processing tasks.

Syntology lines on 16 of the papers shown; no Syntology record for the others (a paper without an arXiv id cannot be joined to the graph, and absence from the graph layer is not a recorded non-run). “Ran” means the sample executed on a synthesized fixture, not that the paper's result was reproduced. Read from the graph 2026-09-24.