Muse Spark 1.3 hits 0.6% attack success rate on Skill-Inject prompt injections
maksym_andr · x · 2026-10-26
Security researcher maksymandr reports that Muse Spark 1.3 achieved only a 0.6% attack success rate on his Skill-Inject benchmark for highly contextual prompt injections, calling it substantial progress. The test was also featured in the model's official system card.
Related event: Muse Spark 1.3 System Card Features Skill-Inject Prompt Injection Test(2 posts)→
More from Models
- Frontier models are better judged by five minutes of weird jokes than 40 benchmark numbers — korymath · 2026-10-26
- Voice models sounding too human called 'the most braindead direction in AI' — yungcontent · 2026-10-26
- Claude free users seemingly can't pick thinking level; UI shows panel but it closes — MilkerGerk · 2026-10-26
- Hamel Husain flags poor benchmark quality: Epoch audits find Terminal-Bench, OSWorld 2.1 all flawed — HamelHusain · 2026-10-26
- Community releases layer-selective abliterated DeepSeek-V4.1 finetune on Hugging Face — AIFlow_ML · 2026-10-26
- User finds Muse's proactive messages surprisingly useful, not annoying — xeophon · 2026-10-26