QQWorld: Quantile-Quantile Matching for World Model Regularization
Latent world models enable efficient planning by predicting future states in a compact representation space, but their performance depends critically on the quality of the learned…
Hugging Face · Daily Papers
·Zhoushun Yu, Xiaoyu Hu
·
·▲ 23 upvotes
Este artigo está em destaque na seleção diária de papers do Hugging Face, curada pela comunidade de pesquisa em IA.
Autores: Zhoushun Yu, Xiaoyu Hu, Xiangyu Xu
- 23 upvotes da comunidade
Resumo
Resumo original (em inglês), extraído do paper:
Latent world models enable efficient planning by predicting future states in a compact representation space, but their performance depends critically on the quality of the learned latent distribution. LeWorldModel (LeWM) regularizes its latents toward an isotropic Gaussian using the Epps-Pulley (EP) objective. We show that the corrective gradients of EP rapidly vanish for isolated tail samples, leaving heavy-tailed deviations insufficiently controlled. To address this limitation, we propose QQWorld, which replaces EP with a quantile-quantile matching objective that directly aligns projected latent samples with rank-matched Gaussian quantiles, thereby maintaining effective corrective gradients in the tails. We further develop cross-batch QQ, which enlarges the effective ranking pool using detached samples from previous batches, and characterize its bias-variance trade-off. Across four control environments, QQWorld effectively improves the average planning success rate of LeWM, while consistently yielding better Gaussian alignment and thinner latent tails.Onde ler
// relacionados
