I wonder if it just learned to prefer least-to-most output style implicitly from RLHF because it leads to better answers.
MACHINE LEARNING
-
Living Systems: Rare Entities Creating Order Through Self-Improvement
By
–
But entities that create order and get better at doing so, aka living things, are rare…
-
AI’s arithmetic: memorization, pattern-matching, tokenization impairs number understanding
By
–
It doesn’t have a calculator. The arithmetic it knows is mostly memorization with pattern-matching and guesswork to fill in the gaps, and there’s too many possible equations to memorize them all. Also, tokenization impairs its lexical understanding of numbers.
-
MLSys Workshop Presentation at NeurIPS 2022
By
–
Thanks for sharing @_akhaliq! Also checkout our oral session and poster at #NeurIPS2022 MLSys workshop! https://
nips.cc/Conferences/20
22/ScheduleMultitrack?event=62290
… -
Tokenization limits: variable tokens, cannot handle per-letter tasks
By
–
Tokenization. It doesn’t see text as sequences of characters, but of variable-length tokens about 4 chars on average. It generally can’t do anything that needs to be done one letter at a time. It’s also bad at counting in its head.
-
AI-Powered Motion Sensors Transform World Cup Ball Tracking
By
–
Hi-Tech World Cup balls charging! For the first time this year, each ball has a motion sensor inside giving real-time ball tracking and statistics. It collects #data such as position, speed and direction, assisting the referees on offside calls Credit: Kinexon
#ml #ai #tech -
ML Demo Hype: Why Cherry-Picked Examples Fool Investors
By
–
Baffled by the number of people (specifically influencers and investors) who still get fooled by ML demos and cherry picked examples engineered for wow effect! Rookie mistake to judge the quality of a model (versus looking at usage, revenue and more tangible evaluations)
-
GPT-3 cannot perform accurate calculations without step-by-step writing.
By
–
In general, you shouldn't expect it to be able to perform accurate calculations, at least not "in its head" — this is a known limitation of models like GPT-3. It only stands a chance if prompted to write calculations out step-by-step like one might do on paper.
-
ChatGPT avoids hallucination on Hofstadter/Bender questions
By
–
You're talking about a different model — this post is about ChatGPT. text‑davinci‑003 still fails on all of the Hofstadter/Bender questions that I've tried. The prompt you're suggesting does not seem to produce a hallucination in ChatGPT: