A PPO-trained agent finds that optimized projective measurements can disentangle small random Clifford circuits with far fewer projections than random measurement rates in MIPT studies.
The sparse reward function is evaluated at the end of every episode
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
quant-ph 1years
2024 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Reinforced Disentanglers on Random Unitary Circuits
A PPO-trained agent finds that optimized projective measurements can disentangle small random Clifford circuits with far fewer projections than random measurement rates in MIPT studies.