PassGPT+: Leveraging Linguistic Priors for Password Modeling
PassGPT+ adapts GPT-2 priors for passwords, recovering 22.53% of RockYou held-out passwords at 10^8 guesses.
PassGPT+ adapts GPT-2's linguistic prior to password data with character-aware tokenization, instead of training from random initialization like PassGAN and PassGPT. On RockYou it recovers 22.53% of held-out passwords within 10^8 guesses, a 16% relative gain over PassGPT, and keeps 79% of that match rate on a disjoint 2020 leak without retraining. The new absorbing-state model PassDiffusion underperforms autoregressive generation by two to three orders of magnitude on exact password matches.
- PassGPT+ adapts GPT-2 priors with character-aware tokenization.
- RockYou recovery is 22.53% at 10^8 guesses, 16% above PassGPT.
- Unretrained transfer to a 2020 leak retains 79% of that match rate.
- PassDiffusion trails autoregressive models by two to three orders of magnitude.
Full article193 words · extracted from arxiv.org · click to collapse
Passwords remain the dominant online authentication mechanism, and understanding how humans choose them is essential for defensive strength estimation and attack simulation alike. Recent learning-based approaches such as PassGAN and PassGPT have shown that deep generative models can learn password structure directly from leaked corpora. However, both train from random initialization on password data alone. The role of linguistic prior knowledge in password modeling, and what it reveals about how humans create secrets, remains largely underexplored. Here, we address this gap with PassGPT+, which adapts the linguistic prior of GPT-2 to password observations through character-aware tokenization. We also introduce PassDiffusion, the first absorbing-state discrete diffusion model for password generation, as a probe of whether non-autoregressive approaches are competitive. On the RockYou benchmark, PassGPT+ recovers 22.53% of held-out passwords at 108 guesses, a 16% relative gain over PassGPT, and retains 79% of this match rate when transferred without retraining to a disjoint 2020 leak dataset, demonstrating that linguistic priors capture persistent regularities of human password generation. PassDiffusion underperforms by two to three orders of magnitude, indicating that autoregressive modeling is substantially better matched than iterative denoising to the discrete, exact-match nature of password generation.
Text extracted automatically; images, tables and formatting may be missing. Original: https://arxiv.org/abs/2609.39880