Beyond Selection: Token Parameterization for Extreme Visual Token Compression

arXiv:2609.35232 · cs.CV, cs.LG · Submitted 2026-09-28 · Read on arXiv

cs.CV, cs.LG

Submitted: 2026-09-28

Updated: 2026-09-29

Comments: Accepted at NeurIPS 2026 (Spotlight). Code: https://github.com/zrrraa/Braco

Code: https://github.com/zrrraa/Braco

License: http://creativecommons.org/licenses/by/4.0/

The gist: Visual-token compression is effective for improving the efficiency of vision-language models, but under extreme compression budgets, token pruning can break visual grounding while learned resamplers

Terminology

Abstract

Visual-token compression is effective for improving the efficiency of vision-language models, but under extreme compression budgets, token pruning can break visual grounding while learned resamplers increase parameter count, attention cost, and training complexity. We revisit compression through a token parameterization lens, separating (i) basis transformation and structured truncation (retained subspace/compressibility) from (ii) coordinate organization (optimization and cross-modal alignment). This view yields two coupled objectives, compressibility and learnability, which we formalize as unified functionals. Guided by these objectives, we design Braco, a lightweight four-step coder that combines transform-basis truncation, input-independent basis-coordinate embeddings, budget-dependent orthogonal re-parameterization, and learned spatial residual tokens from lightweight pooling. Experiments show that Braco forms the favorable empirical accuracy-efficiency frontier under 23 times -- 64 times compression and remains competitive at 144 times, reaching 95.2% accuracy while reducing prefill FLOPs by 84.2%--86.7% relative to the uncompressed upper bound. Against prior methods, Braco matches or improves accuracy while achieving up to approximately 36% end-to-end speedup and using 16.6 times / 78.8 times lower compressor latency/FLOPs.

Sources

Related papers