The 8088 The 8088 ← All news
arXiv cs.CL AI Research Apr 22

Where Fake Citations Are Made: Tracing Field-Level Hallucination to Specific Neurons in LLMs

★★★★★ significance 3/5

Researchers investigated why Large Language Models generate fake citations, finding that author names are particularly prone to hallucination. The study identifies specific 'hallucination neurons' in the Qwen2.5-32B-Instruct model and demonstrates that suppressing these neurons can improve citation accuracy.

Why it matters Identifying specific neurons responsible for hallucinations offers a potential mechanistic pathway for engineering more reliable, fact-based generative outputs.
Read the original at arXiv cs.CL

Entities mentioned

Qwen

Tags

#hallucination #llm #citations #interpretability #qwen

Related coverage