They are statistical text generators, whose results are defined by their training data set. This is why the paper cited reads thusly:
Despite extensive research into the reasoning capabilities
of Large Language Models (LLMs), most studies have failed
to rigorously differentiate between inductive and deductive
reasoning ...
There is no differentiation because what was sought is the existence of what does not.
The authors then postulate:
This raises an essential question: In LLM reasoning, which
poses a greater challenge - deductive or inductive reasoning?
There is no such thing as "LLM reasoning." Therefore, the greatest challenge is accepting this fact and/or that anthropomorphism is a real thing.
They are statistical text generators, whose results are defined by their training data set. This is why the paper cited reads thusly:
There is no differentiation because what was sought is the existence of what does not.The authors then postulate:
There is no such thing as "LLM reasoning." Therefore, the greatest challenge is accepting this fact and/or that anthropomorphism is a real thing.