LLM agents autonomously evolve human-readable safety specifications from sparse 1-bit danger signals, outperforming reward-based reflection that encourages reward hacking.
Inverse reward design
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
citation-role summary
background 1
citation-polarity summary
verdicts
UNVERDICTED 3roles
background 1polarities
background 1representative citing papers
Moral judgments become more deontological when human design of AI is visible, and designers are judged more strictly than the AI or unaided humans, creating plural and non-converging targets for value alignment.
Derives SimPO from MaxEnt RL and reports that MaxEnt RL in online RLHF exhibits frequent overoptimization and unstable KL dynamics across scales, unlike stable KL-constrained baselines.
citing papers explorer
-
Discovering Agentic Safety Specifications from 1-Bit Danger Signals
LLM agents autonomously evolve human-readable safety specifications from sparse 1-bit danger signals, outperforming reward-based reflection that encourages reward hacking.
-
The Alignment Target Problem: Divergent Moral Judgments of Humans, AI Systems, and Their Designers
Moral judgments become more deontological when human design of AI is visible, and designers are judged more strictly than the AI or unaided humans, creating plural and non-converging targets for value alignment.
-
Failure Modes of Maximum Entropy RLHF
Derives SimPO from MaxEnt RL and reports that MaxEnt RL in online RLHF exhibits frequent overoptimization and unstable KL dynamics across scales, unlike stable KL-constrained baselines.