paper-with-me

Papers

TORE: Token Recycling in Vision Transformers for Efficient Active Visual Exploration

2023-11-26 · Jan Olszewski, Dawid Rymarczyk, Piotr Wójcik, Mateusz Pach, Bartosz Zieliński

Active Visual Exploration (AVE) optimizes the utilization of robotic resources in real-world scenarios by sequentially selecting the most informative observations. However, modern methods require a high computational budget due to processing the same observations multiple times through the autoencoder transformers. As a remedy, we introduce a novel approach to AVE called TOken REcycling (TORE). It divides the encoder into extractor and aggregator components. The extractor processes each observation separately, enabling the reuse of tokens passed to the aggregator. Moreover, to further reduce the computations, we decrease the decoder to only one block. Through extensive experiments, we demonstrate that TORE outperforms state-of-the-art methods while reducing computational overhead by up to 90\%.

📄 PDF Abstract BibTeX arXiv:2311.15335

Code (1)

jano1906/TokenRecycling 공식 구현 pytorch

Tasks

Decision MakingDecoderImputationMissing ValuesSequential Decision Making

Similar Papers 제목 키워드 기반

Turning Trash into Treasure: Accelerating Inference of Large Language Models with Token Recycling

2024-08-16 · Xianzhen Luo, YiXuan Wang, Qingfu Zhu, Zhiming Zhang 외

Massive parameters of LLMs have made inference latency a fundamental bottleneck. Speculative decoding represents a lossless approach to accelerate inference through a guess-and-verify paradigm. Some methods rely on addit…

Retrieval

Visual Prompt Tuning for Generative Transfer Learning

2022-10-03 · CVPR 2023 1 · Kihyuk Sohn, Yuan Hao, José Lezama, Luisa Polania 외

Transferring knowledge from an image synthesis model trained on a large dataset is a promising direction for learning generative image models from various domains efficiently. While previous works have studied GAN models…

Image GenerationTransfer LearningVisual Prompt Tuning

Mamba-R: Vision Mamba ALSO Needs Registers

2024-05-23 · Feng Wang, Jiahao Wang, Sucheng Ren, Guoyizhe Wei 외

Similar to Vision Transformers, this paper identifies artifacts also present within the feature maps of Vision Mamba. These artifacts, corresponding to high-norm tokens emerging in low-information background areas of ima…

MambaSemantic Segmentation

Mamba-Reg: Vision Mamba Also Needs Registers

2025-01-01 · CVPR 2025 1 · Feng Wang, Jiahao Wang, Sucheng Ren, Guoyizhe Wei 외

Similar to Vision Transformers, this paper identifies artifacts also present within the feature maps of Vision Mamba. These artifacts, corresponding to high-norm tokens emerging in low-information background areas of…

MambaSemantic Segmentation

RecycleGPT: An Autoregressive Language Model with Recyclable Module

2023-08-07 · Yufan Jiang, Qiaozhi He, Xiaomin Zhuang, Zhihua Wu 외

Existing large language models have to run K times to generate a sequence of K tokens. In this paper, we present RecycleGPT, a generative language model with fast decoding speed by recycling pre-generated model states wi…

Language ModelingLanguage ModellingmodelText Generation