Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
3
0
2
10
Training seeds can dramatically skew recommender system evaluations, revealing that relying on a single seed may misrepresent model performance.
Users can tell the difference in popularity composition of music recommendations, but they don’t necessarily prefer the calibrated options.
Autoguidance—using a model to guide itself—can effectively reduce popularity bias in diffusion recommenders, leading to fairer item exposure without significantly sacrificing accuracy.