Understanding and Enhancing Safety Mechanisms of LLMs via Safety-Specific Neuron

Yiran Zhao, Wenxuan Zhang, Yuxi Xie, Anirudh Goyal, Kenji Kawaguchi +1 more

2/12/2026

No official code found. Best implementation: zhaoyiran924/Safety-Neuron (30 stars, last commit 2025-04-30).

Semantic Scholar

Code Implementations(1)

zhaoyiran924/Safety-NeuronCommunity66%

[ICLR 2025] Understanding and Enhancing Safety Mechanisms of LLMs via Safety-Specific Neuron

302PythonMar 1, 20251 years ago

Cite this paper

@article{zhao2026understanding,
  title  = {Understanding and Enhancing Safety Mechanisms of LLMs via Safety-Specific Neuron},
  author = {Yiran Zhao and Wenxuan Zhang and Yuxi Xie and Anirudh Goyal and Kenji Kawaguchi and Michael Shieh},
  year   = {2026},
  url    = {https://api.semanticscholar.org/CorpusID:f20a924bb7e7a59fd11a36ff2ed9733bc2329d49},
  journal = {ICLR 2025 2025}
}

Understanding and Enhancing Safety Mechanisms of LLMs via Safety-Specific Neuron

Code Implementations(1)

Cite this paper

Discussion