Transformer attention in generative recommenders can create positional, popularity, latent-driver, and synthetic-data biases that concentrate exposure and reduce diversity.
We define wh :=E q(n) |X (n) =h , oa9 which can be interpreted as the direction the query context signal given tokenh
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CY 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
LLM Biases
Transformer attention in generative recommenders can create positional, popularity, latent-driver, and synthetic-data biases that concentrate exposure and reduce diversity.