An offline RL policy (TD3-BC) is claimed to beat Aave's rule-based rates on responsiveness, lender returns, and stress response, but the evidence is limited to historical replay.
In: NeurIPS (2021)
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
REJECT 1representative citing papers
citing papers explorer
-
From Rules to Rewards: Reinforcement Learning for Interest Rate Adjustment in DeFi Lending
An offline RL policy (TD3-BC) is claimed to beat Aave's rule-based rates on responsiveness, lender returns, and stress response, but the evidence is limited to historical replay.