Record summary
A quick snapshot of what this page covers.
Risk profile
How this risk is described and categorized.
"Reward Hacking: In practice, proxy rewards are often easy to optimize and measure, yet they frequently fall shortof capturing the full spectrum of the actual rewards (Pan et al., 2021). This limitation is denoted as misspecifiedrewards. The pursuit of optimization based on such misspecified rewards may lead to a phenomenon knownas reward hacking, wherein agents may appear highly proficient according to specific metrics but fall short whenevaluated against human standards (Amodei et al., 2016; Everitt et al., 2017). The discrepancy between proxyrewards and true rewards often manifests as a sharp phase transition in the reward curve (Ibarz et al., 2018).Furthermore, Skalse et al. (2022) defines the hackability of rewards and provides insights into the fundamentalmechanism of this phase transition, highlighting that the inappropriate simplification of the reward function can bea key factor contributing to reward hacking."
Suggested mitigations
Defenses that may help with related attacks.
Source
Research source for this risk, when available.
Included resource
AI Alignment: A Comprehensive Survey
Original source
MIT AI Risk Repository
Open the public repository used for AI risk records and taxonomy fields.
