DyBBT proposes a cognitive dual-system dialog policy with a bandit-inspired meta-controller that dynamically balances exploration based on real-time states and visitation counts.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2025 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
DyBBT: Dynamic Balance via Bandit-inspired Targeting for Dialog Policy with Cognitive Dual-Systems
DyBBT proposes a cognitive dual-system dialog policy with a bandit-inspired meta-controller that dynamically balances exploration based on real-time states and visitation counts.