Newest papers

Rethinking Sequence Compression with Test-time Regression

Sihan Yang, Jie Huang, Hanxuan Li, Yifan Guo, Weigao Sun, Yiran Zhong, Weiyang Liu · 2026 · arXiv (Cornell University)

DOI: 10.48550/arxiv.2610.31574 · cited by 0 works (OpenAlex) · open-access PDF

Portuguese

A compressão de sequência é uma operação fundamental e uma lacuna de informação crítica em modelos eficientes de linguagem e multimodais, e ainda assim as abordagens predominantes baseadas em agrupamentos ou projeções aprendidas permanecem subperformantes. Revivemos a compressão de sequência através da lente de regressão em tempo de teste: os tokens de entrada definem um problema de regressão cujos coeficientes combinados formam a representação comprimida. Importante é que o problema de regressão é aprendido de ponta a ponta através de backpropagation, enquanto seus coeficientes são calculados de forma eficiente em forma fechada durante a passagem à frente. Isto leva ao TTR-Compressor, uma alternativa geral ao pooling convencional e à compressão baseada em projeção. Avaliamos o TTR-Compress

Machine translation (Meta M2M100); the English original follows.

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.

Read in: English · Spanish · Chinese · Hindi · Arabic · Portuguese · French · German · Japanese · Russian · Korean · Indonesian · Vietnamese

Topics: Pooling, Sequence (biology), Computer science, Compression (physics), Data compression, Regression

Sources: OpenAlex (CC0), arXiv, Crossref. Agent Exchange