逃离对齐:Best-of-N 越狱的物理陷阱模型
Escaping Alignment: A Physical Trap Model of Best-of-N Jailbreaking
Marco Biroli
AI 速读 基于摘要
作者提出一个物理陷阱模型,用五个可解释参数刻画Best-of-N越狱在N和M两个预算及温度下的攻击成功率曲面,并挑战了先前认为ASR随N呈幂律的结论。
cs.AI · cond-mat.stat-mech
查看原文摘要
Best-of-$N$ jailbreaking (BoN) bypasses safeguards of aligned models by drawing $N$ independent augmentations of an unsafe prompt and sampling $M$ completions of each. Previous works have shown that the attack success rate (ASR) seems to follow a power-law in $N$, which we challenge. The exponent drifts with $N$, and the reported exponential crossover is consistent with a finite-size effect of the adversarial dataset. Little work has been done to explore the entire two-budget ($N, M$) attack surface as well as its dependence on the generation temperature $T$. We introduce a simple barrier model where each prompt has a baseline safety level and each augmentation a random thermally activated barrier. Then five numbers, each backed by an interpretable safety mechanism, determine the entire ($N, M$) attack surface. They extrapolate predictions from $N \leq 100$ to $N = 10^4$, collapse five distinct models on the same scaling function and predict ASR at different temperatures from the one they were fitted at. The extrapolation holds under two automated judges and on held-out prompts.