AllocFlow Portfolio Trading Action Generation Framework Released
The story so far
qwen Released AllocFlow Framework, The Framework Generates Joint Portfolio Configurations From Raw Multi-asset Market History Through Cross-asset Flow Transformer and Two-stage Imitation and Group Relative Reinforcement Learning Program, Improving Risk-adjusted Return on Cryptocurrency Perpetual Futures. The Framework Uses Stream Transformer to Generate Continuous, Signed Position Weights for All Tools, and Optimizes Configurations Through Policy Supervised Imitation and Group Relative Reinforcement Learning, Achieving a Calmar Ratio of 4.90 on 6 Months of Unretrained Data in 2026.
AI-written from the reports · Updated 1 hour ago
Coverage timeline
Follow the coverage to see every side of the story.
- Hugging Face Daily PapersAllocFlow: End-to-End Portfolio Trading Action Generation from Raw Market Data
AllocFlow framework generates joint portfolio configurations from raw multi-asset market history through cross-asset flow Transformer and a two-stage imitation and group relative reinforcement learning program, enhancing risk-adjusted returns on cryptocurrency perpetual futures. The framework employs flow Transformer to generate continuous, signed position weights for all instruments and optimizes configurations through policy-supervised imitation and group relative reinforcement learning, achieving a Calmar ratio of 4.90 on 6 months of non-retrained data as of 2026.
Heat over time
Not enough data yet to draw a trend.