---
title: FAST: Efficient Action Tokenization for Vision-Language-Action Models
url: https://www.ml-quant.com/papers/arxiv/2501.09747/
site: ML-Quant (https://www.ml-quant.com)
updated: 2026-09-26
license: Summaries CC BY 4.0; links go to the original sources
index: https://www.ml-quant.com/llms.txt
identifier: arXiv:2501.09747
source_url: https://arxiv.org/abs/2501.09747
featured: 2025-01-23
citations: 673
topic: Trading, Microstructure & Execution
---


# FAST: Efficient Action Tokenization for Vision-Language-Action Models

A new tokenization scheme, Frequency-space Action Sequence Tokenization (FAST), has been proposed for robot actions, facilitating the training of vision-language action policies for complex and high-frequency tasks.

- Source: https://arxiv.org/abs/2501.09747
- Identifier: arXiv:2501.09747
- Released: 2025-01-16
- First featured: Quant Letter No. 83 (2025-01-23): https://www.ml-quant.com/issues/2025-01-23/
- Citations (Semantic Scholar): 673
- Published in: Robotics
- Topic: Trading, Microstructure & Execution

## Related

- [Deep Reinforcement Learning for Active High Frequency Trading](https://www.ml-quant.com/papers/arxiv/2101.07107/): A new Deep Reinforcement Learning framework has been developed for high frequency stock trading, showing potential for profitable long-term strategies.
- [High-frequency lead-lag relationships in the Chinese stock index futures market: tick-by-tick dynamics of calendar spreads](https://www.ml-quant.com/papers/arxiv/2501.03171/): The research examines the relationships between different maturity stock index futures contracts in China's financial futures market, uncovering a regular pattern of price discovery and a profitable trading strategy based on this pattern.
- [Event-Based Limit Order Book Simulation under a Neural Hawkes Process: Application in Market-Making](https://www.ml-quant.com/papers/arxiv/2502.17417/): An event-driven Limit Order Book model using a Neural Hawkes process is proposed to simulate high-frequency dynamics in financial markets, offering a more accurate depiction of trade execution.
- [Optimal Execution under Incomplete Information](https://www.ml-quant.com/papers/arxiv/2411.04616/): The research introduces a high-frequency trading model using a hidden Markov process to examine optimal liquidation strategies under limited information, offering a practical algorithm to simulate the original liquidation issue.
- [High-frequency intraday trading for battery storages](https://www.ml-quant.com/papers/arxiv/2504.06932/): The paper presents an automated trading strategy for battery energy storage systems in the power intraday market, showing increased revenue potential dependent on trading speed.
- [FlowHFT: Imitation Learning via Flow Matching Policy for Optimal High-Frequency Trading under Diverse Market Conditions](https://www.ml-quant.com/papers/arxiv/2505.05784/): High-Frequency Trading: FlowHFT, a new imitation learning framework, outperforms traditional high-frequency trading models by learning from multiple expert models.