HARP: Training-Free Dual-Profile Agentic Communication for LLM-Based Recommendation

In 30 Sekunden

Was passiert ist

HARP adds three operators to frozen LLM recommendation systems: symmetric item profiles, retrieval-level homophily matching, and reflective profile updates without retraining.

Warum es zählt

Engineers building LLM-based recommenders who want better ranking accuracy without modifying the base model or retraining.

Achtung

The method requires access to item metadata for profile extraction and adds computational cost during retrieval; gains vary across datasets.