A sparse autoencoder with synthetic user probes identifies popularity-encoding neurons in a recommender model, and steering those neurons improves exposure fairness with limited accuracy loss.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.IR 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Opening the Black Box: Interpretable Remedies for Popularity Bias in Recommender Systems
A sparse autoencoder with synthetic user probes identifies popularity-encoding neurons in a recommender model, and steering those neurons improves exposure fairness with limited accuracy loss.