Paper Detail

Rethinking Sequence Compression with Test-time Regression

Sihan Yang, Jie Huang, Hanxuan Li, Yifan Guo, Weigao Sun, Yiran Zhong, Weiyang Liu

arxiv Score 11.3

Published 2026-10-22 · First seen 2026-10-23

General AI

Abstract

Sequence compression is a fundamental operation and a critical information bottleneck in efficient language and multimodal models, and yet prevailing approaches based on pooling or learned projections remain underperforming. We revisit sequence compression through the lens of test-time regression: input tokens define a regression problem whose fitted coefficients form the compressed representation. Importantly, the regression problem is learned end-to-end via backpropagation, while its coefficients are computed efficiently in closed form during the forward pass. This leads to TTR-Compressor, a general alternative to conventional pooling and projection-based compression. We evaluate TTR-Compressor as the compressor for block-wise sparse attention and the visual token merger for multimodal large language models. On block-wise sparse attention, TTR-Compressor improves RULER performance over attention-based compressors by a large margin. For visual token merging, it consistently outperforms the standard MLP merger and matches its performance with 55.6% fewer visual tokens. These results establish test-time regression as an effective primitive for sequence compression in efficient language and multimodal models.

Workflow Status

Review status
pending
Role
unreviewed
Read priority
now
Vote
Not set.
Saved
no
Collections
Not filed yet.
Next action
Not filled yet.

Reading Brief

No structured notes yet. Add `summary_sections`, `why_relevant`, `claim_impact`, or `next_action` in `papers.jsonl` to enrich this view.

Why It Surfaced

No ranking explanation is available yet.

Tags

No tags.

BibTeX

@article{yang2026rethinking,
  title = {Rethinking Sequence Compression with Test-time Regression},
  author = {Sihan Yang and Jie Huang and Hanxuan Li and Yifan Guo and Weigao Sun and Yiran Zhong and Weiyang Liu},
  year = {2026},
  abstract = {Sequence compression is a fundamental operation and a critical information bottleneck in efficient language and multimodal models, and yet prevailing approaches based on pooling or learned projections remain underperforming. We revisit sequence compression through the lens of test-time regression: input tokens define a regression problem whose fitted coefficients form the compressed representation. Importantly, the regression problem is learned end-to-end via backpropagation, while its coefficie},
  url = {https://arxiv.org/abs/2610.31574},
  keywords = {cs.LG},
  eprint = {2610.31574},
  archiveprefix = {arXiv},
}

Metadata

{}