Mistral-7B-Instruct-v0.2 + TruthX

Reported on 2 benchmarks across 1 task · 1 paper · 1 SOTA

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Natural Language Processing2 results

Question AnsweringonTruthfulQA
MC2· 2024-02-27
0.75
SOTA
TruthX: Alleviating Hallucinations by Editing Large Language Models in Truthful Space arXiv:2402.17811
Question AnsweringonTruthfulQA
MC1· 2024-02-27
0.56
best: 0.59 (GPT-4 (RLHF))
TruthX: Alleviating Hallucinations by Editing Large Language Models in Truthful Space arXiv:2402.17811