AutoRedTrader generates synthetic financial misinformation via behavioral bias manipulation and agent feedback to red-team LLM trading agents, reaching 69% exposure and 26.67% attack success on Bitcoin data simulations.
Conspemollm-v2: A robust and stable model to detect sentiment-transformed conspiracy theories
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.CE 1years
2026 1verdicts
UNVERDICTED 1roles
background 1polarities
background 1representative citing papers
citing papers explorer
-
AutoRedTrader: Autonomous Red Teaming of Trading Agents through Synthetic Misinformation Injection
AutoRedTrader generates synthetic financial misinformation via behavioral bias manipulation and agent feedback to red-team LLM trading agents, reaching 69% exposure and 26.67% attack success on Bitcoin data simulations.