No article text was recorded for this version.
information-preserving token reduction
Version 1 of 1176 papers257 conceptsfrozen 2026-10-04
Papers in this version
- Blog: Joint Token Pruning and Squeezing Towards More Aggressive Compression of Vision Transformers
- Blog: Energy Minimizing-based token merging for accelerating Transformers
- Blog: ALPINE: An adaptive language-agnostic pruning method for language models for code
- Blog: Spectral Heat Flow for Conservative Token Condensation in Vision-Language Models
- Blog: Multipole Attention for Efficient Long Context Reasoning
- Blog: Efficient Inference of Vision Instruction-Following Models with Elastic Cache
- Blog: FlowMM: Cross-Modal Information Flow Guided KV Cache Merging for Efficient Multimodal Context Inference
- Blog: SPViT: Enabling Faster Vision Transformers via Soft Token Pruning
- Blog: MEDA: Dynamic KV Cache Allocation for Efficient Multimodal Long-Context Inference
- Blog: UHR-BAT: Budget-Aware Token Compression Vision-Language model for Ultra-High-Resolution Remote Sensing
- Blog: CageViT: Convolutional Activation Guided Efficient Vision Transformer
- Blog: FOLDER: Accelerating Multi-Modal Large Language Models with Enhanced Performance
- Blog: Recoverable Compression: A Multimodal Vision Token Recovery Mechanism Guided by Text Information
- Blog: Co-Me: Confidence-Guided Token Merging for Visual Geometric Transformers
- Blog: iLLaVA: An Image is Worth Fewer Than 1/3 Input Tokensin Large Multimodal Models
- Blog: Efficient Inference of Vision Instruction-Following Models with Elastic Cache
- Blog: LOOK-M: Look-Once Optimization in KV Cache for Efficient Multimodal Long-Context Inference
- Blog: SPViT: Enabling Faster Vision Transformers via Soft Token Pruning
- Blog: A General and Efficient Training for Transformer via Token Expansion
- Blog: Causal-Inspired Multitask Learning for Video-Based Human Pose Estimation
- Blog: Learning Soft Sparse Shapes for Efficient Time-Series Classification
- Blog: DTMFormer: Dynamic Token Merging for Boosting Transformer-Based Medical Image Segmentation
- Blog: Vision Langauge Pre-training by Contrastive Learning with Cross-Modal Similarity Regulation
- Blog: HoliTom: Holistic Token Merging for Fast Video Large Language Models
- Blog: Less is More: A Simple yet Effective Token Reduction Method for Efficient Multi-modal LLMs
- Blog: Token Merging for Fast Stable Diffusion
- Blog: Beyond Attentive Tokens: Incorporating Token Importance and Diversity for Efficient Vision Transformers
- Blog: D2O: Dynamic Discriminative Operations for Efficient Long-Context Inference of Large Language Models
- Blog: SAT-HMR: Real-Time Multi-Person 3D Mesh Estimation via Scale-Adaptive Tokens
- Blog: A General and Efficient Training for Transformer via Token Expansion
- Blog: DocKylin: A Large Multimodal Model for Visual Document Understanding with Efficient Visual Slimming
- Blog: Recoverable Compression: A Multimodal Vision Token Recovery Mechanism Guided by Text Information
- Blog: Efficient Long-Context LLM Inference via KV Cache Clustering
- Blog: Joint Token Pruning and Squeezing Towards More Aggressive Compression of Vision Transformers
- Blog: Self-slimmed Vision Transformer
- Blog: Making Vision Transformers Efficient from A Token Sparsification View
- Blog: Block-based Symmetric Pruning and Fusion for Efficient Vision Transformers
- Blog: Rethinking Visual Token Reduction in LVLMs under Cross-modal Misalignment
- Blog: DocKylin: A Large Multimodal Model for Visual Document Understanding with Efficient Visual Slimming
- Blog: LOOK-M: Look-Once Optimization in KV Cache for Efficient Multimodal Long-Context Inference
- Blog: iLLaVA: An Image is Worth Fewer Than 1/3 Input Tokens in Large Multimodal Models
- Blog: Efficient Vision-and-Language Pre-training with Text-Relevant Image Patch Selection
- Blog: Less is More: A Simple yet Effective Token Reduction Method for Efficient Multi-modal LLMs
- Blog: VLM-Pruner: Buffering for Spatial Sparsity in an Efficient VLM Centrifugal Token Pruning Paradigm
- Blog: Token-Sparse Medical Multimodal Reasoning via Dual-Stream Reinforcement Learning
- Blog: Self-slimmed Vision Transformer
- Blog: Video, How Do Your Tokens Merge?
- Blog: SToRM: Supervised Token Reduction for Multi-modal LLMs toward efficient end-to-end autonomous driving
- Blog: PACT: Pruning and Clustering-Based Token Reduction for Faster Visual Language Models
- Blog: MEDA: Dynamic KV Cache Allocation for Efficient Multimodal Long-Context Inference
- Blog: Make Your ViT-based Multi-view 3D Detectors Faster via Token Compression
- Blog: CrossGET: Cross-Guided Ensemble of Tokens for Accelerating Vision-Language Transformers
- Blog: ELIP: Efficient Language-Image Pre-training with Fewer Vision Tokens
- Blog: FOLDER: Accelerating Multi-modal Large Language Models with Enhanced Performance
- Blog: SparseVLM: Visual Token Sparsification for Efficient Vision-Language Model Inference
- Blog: CrossGET: Cross-Guided Ensemble of Tokens for Accelerating Vision-Language Transformers
- Blog: ToFu: Visual Tokens Reduction via Fusion for Multi-modal, Multi-patch, Multi-image Task
- Blog: WeightedKV: Attention Scores Weighted Key-Value Cache Merging for Large Language Models
- Blog: D2O: Dynamic Discriminative Operations for Efficient Generative Inference of Large Language Models
- Blog: Token-level Correlation-guided Compression for Efficient Multimodal Document Understanding
Showing 60 of 176 papers.
History & citation
Current · Permanent link
- Context-Preserving Token Reduction in Multimodal Transformers merged into this direction (75% overlap)
- Semantic Token Aggregation for Efficient Multimodal Transformers merged into this direction (71% overlap)
- Context-Preserving Token Merging for Efficient Transformers merged into this direction (100% overlap)
Cite this direction
Latest version (a living page):
@misc{lacuna_information_preserving_token_reduction_052488b9,
title = {information-preserving token reduction},
author = {{Lacuna}},
howpublished = {Lacuna research direction},
year = {2026},
url = {https://lacuna.tiptreesystems.com/direction/information-preserving-token-reduction/rdr_052488b9f063484f9e554954d85fa732},
note = {Latest version of a living research direction; cite a version for a fixed text. Accessed 2026-11-20},
}This exact version (version 1, frozen text):
@misc{lacuna_information_preserving_token_reduction_052488b9_v1,
title = {information-preserving token reduction},
author = {{Lacuna}},
howpublished = {Lacuna research direction},
year = {2026},
url = {https://lacuna.tiptreesystems.com/direction/information-preserving-token-reduction/rdr_052488b9f063484f9e554954d85fa732/v/1},
note = {Version 1, frozen 2026-10-04. Accessed 2026-11-20},
}