Echoes in the Loop: Diagnosing Risks in LLM-Powered Recommender Systems under Feedback Loops
Fuente:
arXiv
Salvato in:
| Autori principali: | , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2026
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866918326760898560 |
|---|---|
| author | Park, Donguk Lee, Dongwon Lee, Yeon-Chang |
| author_facet | Park, Donguk Lee, Dongwon Lee, Yeon-Chang |
| contents | Large language models (LLMs) are increasingly embedded into recommender systems, where they operate across multiple functional roles such as data augmentation, profiling, and decision making. While prior work emphasizes recommendation performance, the systemic risks of LLMs, such as bias and hallucination, and their propagation through feedback loops remain largely unexplored. In this paper, we propose a role-aware, phase-wise diagnostic framework that traces how these risks emerge, manifest in ranking outcomes, and accumulate over repeated recommendation cycles. We formalize a controlled feedback-loop pipeline that simulates long-term interaction dynamics and enables empirical measurement of risks at the LLM-generated content, ranking, and ecosystem levels. Experiments on widely used benchmarks demonstrate that LLM-based components can amplify popularity bias, introduce spurious signals through hallucination, and lead to polarized and self-reinforcing exposure patterns over time. We plan to release our framework as an open-source toolkit to facilitate systematic risk analysis across diverse LLM-powered recommender systems. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2602_07442 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Echoes in the Loop: Diagnosing Risks in LLM-Powered Recommender Systems under Feedback Loops Park, Donguk Lee, Dongwon Lee, Yeon-Chang Human-Computer Interaction Information Retrieval Large language models (LLMs) are increasingly embedded into recommender systems, where they operate across multiple functional roles such as data augmentation, profiling, and decision making. While prior work emphasizes recommendation performance, the systemic risks of LLMs, such as bias and hallucination, and their propagation through feedback loops remain largely unexplored. In this paper, we propose a role-aware, phase-wise diagnostic framework that traces how these risks emerge, manifest in ranking outcomes, and accumulate over repeated recommendation cycles. We formalize a controlled feedback-loop pipeline that simulates long-term interaction dynamics and enables empirical measurement of risks at the LLM-generated content, ranking, and ecosystem levels. Experiments on widely used benchmarks demonstrate that LLM-based components can amplify popularity bias, introduce spurious signals through hallucination, and lead to polarized and self-reinforcing exposure patterns over time. We plan to release our framework as an open-source toolkit to facilitate systematic risk analysis across diverse LLM-powered recommender systems. |
| title | Echoes in the Loop: Diagnosing Risks in LLM-Powered Recommender Systems under Feedback Loops |
| topic | Human-Computer Interaction Information Retrieval |
| url | https://arxiv.org/abs/2602.07442 |