SEP 9, 2026 · PREPRINT
Forgetting Only What Matters: Layer-Selective Unlearning toward Robust LLMs
arXiv
First-in-human algorithmic method for selective layer unlearning in LLMs, evaluated on benchmarks but without peer review, formal guarantees, or clinical/real-world deployment evidence.
Study details
DesignAlgorithmic development with empirical benc…
InterventionFOM-UL: layer-selective unlearning framewor…
ComparatorSix baselines: gradient ascent (GA), negati…