CodeShrink: Adaptive Visual Compression for Efficient Multimodal Code Understanding 文章

ArXiv CS.CV2026-08-03PAPERen作者: Wenxin Tang, Jingyu Xiao, Zhenyu Liu, Zipeng Xie, Junliang Liu, Wang Luo, Yuan Jiang, Yintong Huo, Michael Lyu
大语言模型
查看原文 →
关系图谱

CodeShrink: Adaptive Visual Compression for Efficient Multimodal Code Understanding · 相关技术

相关技术

DAPTDINDESDELACROSACTAbstractAgentACE3D LUTABIAHATANNARGARIASKSATEAdaptive Batch ScalingAdvanced Persistent ThreatsAmodal Mask Prediction and ROI Re-Alignment (AMPR)Audio Mamba EncoderAugmented Latent IntrinsicsAutoregressive large language modelsAutoregressive multimodal large language modelsBALANCEBidirectional Evolutionary SearchBlurry Window AttentionCEMCOMCOSECRECROPCTICTUCalibrated Entropy ScoreCamouflaged object detectionCardiology Interface TerminologyCartridges at Scale (CAS)Confidence-Guided Early StoppingDABC