20 results for “crowding-out effect”
CS papers onlyHybrid search: Keyword + semantic, ranked by combined score.ⓘ
Want pure semantic search? Try claim verification →
This study examines the effects of combining digital nudges on decision-making behavior.
This paper analyzes multi-model self-consuming training, showing that while human curation helps individual models, cross-model interactions can degrade long-term alignment by dampening or inverting t…
Weiqin Yang, Yue Pan, Chongming Gao, Sheng Zhou +3 more
This paper addresses the issue of popularity bias in transformer-based sequential recommenders by proposing a regularization method to mitigate spectral collapse.
This paper analyzes interaction data from touch-enabled exhibits at a science center to derive a behavioral signature called interaction density and uses it to distinguish between fast-paced games and…
This paper shows that in the large depth and width asymptotics, Dropout and Random Gradient Masking (RaM) converge to the same limiting dynamics for ResNets.
Jiyuan Yang, Gengxin Sun, Mengqi Zhang, Lingjie Wang +4 more
Generative recommenders reduce information cocoons in generative recommendation
This paper studies how popularity-biased training affects the user geometry in collaborative filtering embeddings using the mean-centered user covariance, and proves that it converges to a noise floor…
The paper introduces Diversity-inducing Initialization (DivIn), a novel method that improves image diversity by re-weighting the initial noise selection based on the guidance potential, thereby mitiga…
This paper introduces Neural Expansion, a framework that explains deep neural network generalization through the expansion of directionally-selective ray coverage in the input space.
This paper predicts the aggregate crowd salience of a document from its text before its marks accumulate.
This paper explores the relationship between noise prediction and denoising in finite diffusion networks and proposes a method to mitigate the noise-transmission burden.
This paper proposes Popularity-Aware Denoising (PAD), a framework to improve denoising recommendation methods by modulating denoising strength based on item popularity.
This paper introduces a variance-weighted batch distribution to improve the diversity of samples in diffusion models.
The paper proposes a privacy-preserving system for crowd monitoring that counts individuals across different locations and time periods using face recognition without ever revealing personal identitie…
This paper explores the possibility of using intrinsic device noise in analog neuromorphic hardware as a consolidation mechanism instead of an accuracy tax.