Any arxiv paper or report?

[amin2718]

We have a detailed write-up on the methodology employed here: Cut the Bull…. Detecting Hallucinations in Large Language Models. Simon, the author, was also the lead researcher who carried out development of the model.

[TomLucidor]

Here is the question: How do you manage Qwen or Llama or DeepSeek or Gemma to halloucinate less and be on par with API models?

P.S. GLM is what CodeGeeX is based on, which is very surprising considering it has lower risk of halloucination.