USTC and Alibaba's CEDAR cuts demand forecast error 57% by decoupling decisions from external shocks

量子位 · wechat · 2026-09-17

CEDAR (KDD'26, USTC + Alibaba) models "decision-conditioned simulation": an Action-Interleaved Transformer treats states and actions as first-class tokens, while a residual module uses LLM-extracted social-media signals with cross-attention to capture external shocks. On 32M product trajectories from 1688 it cuts forecast error 57%+ vs top baselines; online A/B tests lifted merchant LTV 13% and ROI 15%.

Original post →

More from Research

Research channel →