Single Biased Sample Can Break Alignment
radamihalcea · x · 2026-07-10
This paper points out that a single biased example can compromise an already aligned LLM. The authors state that training with single-shot GRPO on a flipped label can induce systematic bias that generalizes across attributes, categories, and benchmarks.
More from Research
- Could 10k agents discover learning methods beyond backprop, or just tweak existing ones? — SeunghyunSEO7 · 2026-09-11
- Cognition's SWE-2 uses a KKT duality argument in RL to shift the effort Pareto curve — YouJiacheng · 2026-09-11
- VidMap uses RoMa coarse matching on all frames, fine-scale only for keyframes — ducha_aiki · 2026-09-11
- Bug Hunt Bench author: leaderboard noise is about 2-3 points — PawelHuryn · 2026-09-11
- PNAS paper shows a tiny billiard-ball system is a universal computer — undecidability lives in two dimensions — eigensteve · 2026-09-11
- New paper: Absolute pose estimation from affine cues and gravity direction — ducha_aiki · 2026-09-11