LLM rerankers in cold-start recsys show recall@200 of 0.109, concentrate on only 3 items, and are beaten by popularity baselines (HR@10 0.268 vs 0.008).
ColdRAG: Retrieval-Augmented Generation for Cold- Start Recommendation
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
years
2026 2representative citing papers
Introduces quantile-robust scaling and learned channel scales for SmoothRot transforms, reporting up to 18.5% error reduction on LLaMA-3.2-1B under W4A4 quantization.
citing papers explorer
-
Diagnosing LLM-based Rerankers in Cold-Start Recommender Systems: Coverage, Exposure and Practical Mitigations
LLM rerankers in cold-start recsys show recall@200 of 0.109, concentrate on only 3 items, and are beaten by popularity baselines (HR@10 0.268 vs 0.008).