Differential TD learning converges with local clocks but diverges with global clocks in average-reward RL.
A note on a conjecture concerning rank one perturbations of singular m-matrices
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
On the Divergence of Differential Temporal Difference Learning without Local Clocks
Differential TD learning converges with local clocks but diverges with global clocks in average-reward RL.