Book
Open access
Jul 2026
TRACE: Term-level Reasoning And Chain-of-thought Enhanced distillation for E-commerce Multi-modal Relevance Learning
The proposed TRACE (Term-level Reasoning And Chain-of-thought Enhanced distillation), a framework designed for deploying advanced reasoning capabilities at scale, enhances an LLM's multi-modal reasoning by employing Group Relative Policy Optimization (GRPO) guided by a term-level Chain-of-Thought (CoT) reward function, enabling it to generate detailed, step-by-step relevance judgments.
Chenxu Wang, Chi Zhang, Fang liang et al.
· Annual International ACM SIG... · 0 citations