A latent-space extension of MBPO for measuring and mitigating the sim-to-real gap performs inconsistently across MuJoCo perturbations, and its gap metric is non-monotonic.
Policy invariance under reward transformations: Theory and application to reward shaping
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.LG 1years
2025 1verdicts
REJECT 1roles
background 1polarities
support 1representative citing papers
citing papers explorer
-
Revealing the Challenges of Sim-to-Real Transfer in Model-Based Reinforcement Learning via Latent Space Modeling
A latent-space extension of MBPO for measuring and mitigating the sim-to-real gap performs inconsistently across MuJoCo perturbations, and its gap metric is non-monotonic.