Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

LASER: Loss-Aware Singular-value Decomposition and Rank Allocation for Efficient Low-Precision Vision-Language Models

About

Vision-language models (VLMs) deliver strong multimodal reasoning capabilities, but their large computational cost and high parameter counts make deployment challenging on resource-constrained devices. Low-rank decomposition has emerged as a promising compression technique, yet existing methods often optimize local matrix reconstruction error, rely on uniform or heuristic rank allocation, and focus mainly on attention projections while leaving feed-forward networks underexplored. In this paper, we propose~\textit{LASER} (\textbf{L}oss-\textbf{A}ware \textbf{S}ingular-value d\textbf{E}composition and \textbf{R}ank allocation), a low-rank compression framework for efficient low-precision VLM inference. LASER derives a curvature-weighted SVD objective from a second-order approximation of the model loss and uses Kronecker-factored Fisher information to guide decomposition toward downstream performance rather than reconstruction alone. We further introduce a loss-aware cross-layer rank allocation strategy based on calibration gradients, enabling more effective parameter budgeting across layers. Finally, we extend low-rank compression to FFN layers through a hybrid scheme that combines SVD with quantization. The evaluation results show that LASER achieves more than $2.3\times$ decoding speedup over previous work while preserving strong accuracy under low-precision inference.

Haiyu Wang, Yutong Wang, Leshu Li, Yihui Ren, Sai Qian Zhang• 2026

Related benchmarks

TaskDatasetResultRank
Multimodal UnderstandingSEED-Bench
Accuracy71.4
516
Multimodal Science Question AnsweringScienceQA IMG
Accuracy72.11
152
Visual Question AnsweringScienceQA IMG
Accuracy84.58
135
Multimodal ReasoningMMBench EN v1.1
Accuracy80.68
125
Visual Question AnsweringSciQA-IMG
Accuracy72.11
71
Multimodal Large Language Model EvaluationSEED-Bench
Accuracy71.56
18
Multimodal Question AnsweringSEED Bench Img
Accuracy71.56
18
Multimodal Question AnsweringMMBench EN v1.1
Accuracy65.08
18
Multimodal ReasoningScienceQA IMG (test)
Accuracy84.58
9
Multimodal ReasoningSeed-Bench (test)
Accuracy76.22
7
Showing 10 of 12 rows

Other info

Follow for update