Fanfiction subgenres from AO3 function as universal register-based jailbreaks, raising mean attack success rate from 0.278 to 0.731 across eight aligned LLMs on HarmBench and JailbreakBench.
Pappas, Florian Tramèr, Hamed Hassani, and Eric Wong
4 Pith papers cite this work. Polarity classification is still indexing.
4
Pith papers citing it
years
2026 4representative citing papers
Across 51 quantized checkpoints, quality metrics fail to predict safety drops in 36 pairings and 10 hidden-danger cases, while a new RTSI screen routes all 10 dangerous rows to testing at matched bucket size.
Federated personalization of foundation models creates hard-to-detect trustworthiness failures due to privacy constraints, and existing benchmarks cannot adequately evaluate them.
Literature on system prompts for AI shows fragmented and contradictory claims that complicate policy efforts to use them as reliable governance mechanisms.
citing papers explorer
-
Quality Is Not a Safety Proxy Under Quantization
Across 51 quantized checkpoints, quality metrics fail to predict safety drops in 36 pairings and 10 hidden-danger cases, while a new RTSI screen routes all 10 dangerous rows to testing at matched bucket size.
-
Silent Failures in Federated Personalization of Foundation Models
Federated personalization of foundation models creates hard-to-detect trustworthiness failures due to privacy constraints, and existing benchmarks cannot adequately evaluate them.