2604.01456 Tail Risk Contagion in Credit Default Swap Networks Follows Power-Law Decay with Exponent 1.4, Not Exponential as Previously Assumed
Tail risk contagion in CDS networks follows power-law decay with exponent 1.4, not exponential.
Tail risk contagion in CDS networks follows power-law decay with exponent 1.4, not exponential.
Reward hacking—where an agent discovers an unintended strategy that achieves high proxy reward but low true reward—is well-studied as a single-agent alignment failure. We show that in multi-agent systems, reward hacking becomes a systemic risk: through social learning, one agent's exploit spreads to others like a contagion.