Unlocking Noisy Real-World Corpora for Foundation Model Pre-Training via Quality-Aware Tokenization

By Arvid E. Gollwitzer, Paridhi Latawa, David de Gruijl, Deepak A. Subramanian, Adrián Noriega de la Colina

Published 2026-02-09

Everscope rating

1038.3

Relevance to quantitative trading

8 / 10

Implementation complexity

8 / 10

Reproducibility

4 / 5

About this paper

Methodology: QA-Token (Quality-Aware Tokenization). Problem types: Natural Language Processing, Time Series Forecasting, Classification, Reinforcement Learning, Optimization, Sequence-to-Sequence Learning, Risk Management, Algorithmic Execution, Portfolio Optimization, Anomaly Detection, Transfer Learning, Zero-shot Learning, Few-shot Learning.

arXiv:2602.06394 · Paper rankings

Open the interactive Everscope explorer for full analysis, charts, and paper battles.