More is more: Addition bias in large language models
(sciencedirect.com)
In this paper, we investigate the presence of addition bias in Large Language Models (LLMs), drawing a parallel to the cognitive bias observed in humans where individuals tend to favor additive over sub-tractive changes [3].
In this paper, we investigate the presence of addition bias in Large Language Models (LLMs), drawing a parallel to the cognitive bias observed in humans where individuals tend to favor additive over sub-tractive changes [3].