Gary Marcus

Rank 28 of 47
|
Score -2
6/27/2024, 3:50:14 AM
X
In reply to:
Philosophy Bear
@sumdepony
·
127d
@GaryMarcus Many of the closed weight models are, however, far above chance
Gary Marcus
@GaryMarcus
·
127d
Wondering how LLMs do on the Comprehension Challenge I proposed in 2014?

New results from an easier (written not visual) version of that test:

“no open-weight model performs above random chance (despite their strong performance on synthetic benchmarks)”

https://arxiv.org/pdf/2406.16264

The statement '@sumdepony But a long way from a high school student, I would imagine.' is a lighthearted, conversational remark in response to a discussion about the performance of language models. It does not substantively engage with public issues, policies, or societal concerns, and therefore does not constitute public discourse.