Almanac
paper

Beyond Uniform Tokens: Adaptive Compression for Time Series Language Models

paperactiveprovisionalbeyond-uniform-tokens-adaptive-compression-for-time-series-language-models-6efc6599·1 events·first seen 5d ago

Aliases: Beyond Uniform Tokens: Adaptive Compression for Time Series Language Models

More like this (12)

Recent events (1)

5arXiv · cs.CL·5d ago·source ↗

Adaptive asymmetric token compression accelerates time series language models up to 7.68×

A new arXiv preprint proposes an adaptive token budgeting framework for time series (TS) language models that compresses TS tokens using frequency-domain structure and progressively prunes prompt tokens across model layers. The authors demonstrate up to 7.68× inference acceleration with performance improvements in 78% of evaluated settings across forecasting, classification, imputation, and anomaly detection tasks. The work is motivated by the observation that TS tokens have uneven spectral contributions and prompt-token influence attenuates with model depth, making uniform token processing wasteful.