Skip to content

Translation in progress Some entries still include text in the original language.

#多模态

Indexed today 131 items

Page 11 Newer contentBack to latest

Oct 1Thu
  1. arXiv:对齐、欺骗与监督 · Indexed · Original Paper35

    CoRe: Co-evolving reward models to mitigate latent reward hacking in video diffusion models(original, opens in a new tab)

    To address the problem that optimizing against a fixed latent reward model leads to latent reward hacking and degradation of perceptual and motion quality in generated videos, the researchers propose the co-evolving reward framework CoRe, which treats latent-space alignment as a dynamic interaction between the generator and the reward model, continuously refitting the reward model with the generator's current samples and anchoring to real video preferences, so that the generator cannot obtain high rewards by deviating from the data distribution.