Rlhf on Neputer Bloghttps://blog.neputer.com/tags/rlhf/Recent content in Rlhf on Neputer BlogHugoen-usSun, 18 Oct 2026 13:00:12 +0545New 'GREAT' Backdoor Attack Targets RLHF with Emotion-Aware Triggershttps://blog.neputer.com/news/2026-10-18-new-great-backdoor-attack-targets-rlhf-with-emotion-awa/Sun, 18 Oct 2026 13:00:12 +0545https://blog.neputer.com/news/2026-10-18-new-great-backdoor-attack-targets-rlhf-with-emotion-awa/Researchers reveal GREAT, an RLHF backdoor attack using emotionally angry triggers to hijack models for specific subpopulations, bypassing existing defenses.