Gary Marcus

Rank 27 of 47
|
Score 26
7/19/2024, 2:17:22 PM
X
In reply to:
Nikhil Sulghur
@Nikhilsulghur
·
99d
@GaryMarcus @stephen_wolfram as far as i see it, the syntax part of the table was trained through massive amounts of data and the functionality part needs more of a symbolic+inference improvement. would you say that's correct?
Gary Marcus
@GaryMarcus
·
99d
Incredible. Two years of near-constant overuse of the word “exponential” and on this new @stephen_wolfram benchmark, GPT-4 still comes out on top.

Sure looks like diminishing returns to me.
Stephen Wolfram
@stephen_wolfram
·
100d
How do the LLMs compare? Leveraging our "code grading" tech to introduce weekly computationally grounded LLM benchmarking...
https://www.wolfram.com/llm-benc…

The statement is part of a technical discussion about the performance of large language models (LLMs) and their benchmarking. The tone is inquisitive and seeks clarification from an expert in the field. The intent is to engage in a constructive dialogue about the technical aspects of LLMs, specifically their syntax and functionality. The content is focused on understanding and improving the performance of these models.

  1. Principle 1:
    I will strive to do no harm with my words and actions.
    The statement strives to do no harm and is focused on technical aspects rather than personal attacks or harmful rhetoric. [+1]
  2. Principle 3:
    I will use my words and actions to promote understanding, empathy, and compassion.
    The statement promotes understanding and empathy by seeking clarification and engaging in a constructive dialogue. [+1]
  3. Principle 4:
    I will engage in constructive criticism and dialogue with those in disagreement and will not engage in personal attacks or ad hominem arguments.
    The statement engages in constructive criticism and dialogue, asking for expert opinion to validate a technical hypothesis. [+1]