Recommender Systems

Where does LLM recommendation bias actually come from?

Do conversational AI systems inherit popularity bias from their training data or from the datasets they're deployed on? Understanding the source matters for knowing how to fix it.

Note · 2026-05-03 · sourced from Recommenders Conversational
What breaks when specialized AI models reach real users? Why do LLMs fail at understanding what remains unsaid?

When GPT-4 recommends in conversational-recommendation benchmarks, the most-frequently-recommended items are not the most-popular items in the dataset. They are the most-popular items in some external distribution — presumably the LLM's pretraining corpus.

Empirically: on ReDIAL, popular movies in ground truth like "Avengers: Infinity War" appear about 2% of the time. On Reddit-Movie, popular ground-truth movies like "Everything Everywhere All at Once" appear less than 0.3%. But GPT-4's recommendations concentrate on different items: "The Shawshank Redemption" appears around 5% on ReDIAL and 1.5% on Reddit. The same kinds of items dominate across datasets even though the datasets have different population biases.

This is a different kind of popularity bias than the one collaborative filtering produces. CF popularity bias amplifies the most-clicked items in your training data; the LLM bias imports popularity from a corpus the LLM saw before any of this data existed. It cannot be debiased by the usual dataset-level correction methods because the bias source isn't in the dataset.

The risk is bias-amplification loops: LLM CRS deployed in a recommendation product trains future user behavior on its biased outputs, which shifts the dataset toward LLM-pretraining-popular items, which next-generation LLMs ingest, which deepens the concentration. Different datasets that should produce different recommendations converge on the same set of "canonical popular items" inherited from the web's general distribution.

The implication for production systems: pretraining-corpus popularity is a domain-shift effect that LLM-as-CRS inherits by construction. Mitigating it requires either dataset-aware fine-tuning or post-hoc re-ranking by dataset-specific popularity priors — and probably both.


Source: Recommenders Conversational

Related concepts in this collection

Concept map
13 direct connections · 85 in 2-hop network ·medium cluster

Click a node to walk · click center to open · click Open full network for a force-directed map

your link semantically near linked from elsewhere
Original note title

LLM CRS recommendations exhibit popularity bias inherited from pretraining corpus not from target dataset