According to Cointelegraph, scientists from the University of Science and Technology of China and Tencent’s YouTu Lab developed a tool to combat “hallucination” by artificial intelligence (AI) models.
This subset of AI is expected to include GPT-4 (especially its visual variant, GPT-4V) and other systems that roll vision and/or other processing into the generative AI modality alongside text-based language modeling, Cointelegraph added.
Sources revealed that the new model include GPT-3.5 turbo, Grounding DINO and BLIP-2-FlanT5. Together, these models is believed to work as evaluators to identify hallucinations and instruct the model being corrected to regenerate its output in accordance with its data.
Furthermore, “They evaluated numerous “off the shelf” MLLMs using their method and concluded that Woodpecker could be “easily integrated into other MLLMs,” Cointelegraph concluded.
(With insights from Cointelegraph)