Dual static CVaR decompositions suffer from a CVaR evaluation gap caused by empty intersections in risk-assignment consistency constraints, and no single policy can be optimal across all risk levels in some MDPs.
Czarnecki, Micha \"e l Mathieu, Andrew Dudzik, Junyoung Chung, David H
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
On the Fundamental Limitations of Dual Static CVaR Decompositions in Markov Decision Processes
Dual static CVaR decompositions suffer from a CVaR evaluation gap caused by empty intersections in risk-assignment consistency constraints, and no single policy can be optimal across all risk levels in some MDPs.