Researchers tackle AI fact-checking failures with new LLM training technique ByBrayan Zanelatto August 16, 2024