Financial Context-Rich Strategy for Actor-Critic based Agents: A Deep Reinforcement Learning Approach for Stock Trading
Resumo indisponível. Consulte o paper original.
PAPER / ARXIV:2610.26662
Tian Zhang, Meng Li
RESUMO
LLM trading agents are evaluated almost exclusively in backtests and autonomous benchmark harnesses. Deployment, where actions are irreversible, raises a different question: who verifies the agent's actions, and which failures can each verifier structurally catch? We report on AllPath, a personal LLM trading agent deployed against a live paper-brokerage account. Verification is split across three layers: code verifies what must not depend on judgment (deterministic gates enforce risk limits), the agent verifies itself within fixed bounds (nightly review with an iteration cap and no acting tools), and a human verifies judgment through approval screens. Each layer caught a different kind of failure; the most instructive production incidents passed every gate and were found by hand, with the fix each time being an interface change that makes the condition visible.
NO MESMO MAPA
Resumo indisponível. Consulte o paper original.
Resumo indisponível. Consulte o paper original.
Resumo indisponível. Consulte o paper original.
Resumo indisponível. Consulte o paper original.