I think what Min et al. shows is that much of the benefit of examples is learning the output syntax and label space. Results aren’t as clean as this anecdote suggests by itself, e.g. OOD labels dramatically hurt performance.
MACHINE LEARNING
-
Few-shot Prompt Setup Notes for Scaling Law Experiments
By
–
Some setup notes (cc @EthanJPerez
) – We used the exact 2-shot prompt for Quote Repetition, which is already U-shaped for Gopher/Chinchilla – We used fewer shots for Hindsight – We did few-shot instead of 0-shot for Negation QA – We also showed inverse scaling up to PaLM 62B -
U-shaped Scaling and the Limitations of Inverse Scaling Tasks
By
–
Implications: 1. U-shaped scaling means that inverse scaling may not hold when extrapolated to larger models. 2. The term “inverse scaling task” is underspecified. A task can be inverse scaling for one type of prompting and positive or U-shaped for another type of prompting.
-
CoT Prompting Defends Against Inverse Scaling in Math Tasks
By
–
Second, we show that CoT prompting can defend against inverse scaling. For instance, CoT prompting achieves 100% on 7 out of 8 subtasks for Redefine Math.
-
U-shaped Scaling Behavior Emerges at Higher Computational Budgets
By
–
Our results first confirm inverse scaling behavior seen on prior models trained up to 500 zettaFLOPs. But at 2K zettaFLOPs, it becomes U-shaped. U-scaling has also been shown in prior work, such as BIG-Bench.
-
Inverse Scaling Becomes U-Shaped with Larger Language Models
By
–
New preprint!
— Jason Wei (@_jasonwei) 4 novembre 2022
By evaluating 5x larger language models, inverse scaling can become “U-shaped scaling”, which means that performance increases sharply after decreasing.
https://t.co/bZQndKqlB6
These two tasks here are Third Prize winners from the Inverse Scaling Prize. pic.twitter.com/8d3pu8DDrkNew preprint! By evaluating 5x larger language models, inverse scaling can become “U-shaped scaling”, which means that performance increases sharply after decreasing. https://
arxiv.org/abs/2211.02011 These two tasks here are Third Prize winners from the Inverse Scaling Prize. -
JMLR Review Delays: How TMLR Improves Academic Publishing
By
–
I had the same experience with JMLR back in 2015. After 6 months, they hadn't found reviewers. Withdrew the paper and submitted to a different journal. I like that @TmlrOrg doesn't run into this type of friction, since reviewers are expected to review it after being assigned.
-
Meta AI Creates Largest Protein Language Model with 15B Parameters
By
–
Meta AI researchers trained a language model to fill in protein sequence gaps across millions of diverse proteins & scaled up to 15B parameters, creating the largest language model of proteins to date. More on our latest breakthrough in protein folding https://
bit.ly/3WoWcK2 -
Machine Learning GPT3 Website Free for Content Creators
By
–
Machine learning for deep thought: a GPT3-based website that every content creator should jump on while it’s free… https://
qburst.co -
ERNIE-Layout Achieves SOTA Results on Document Understanding Tasks
By
–
Introducing ERNIE-Layout, a layout knowledge enhanced document pre-training model that obtained SOTA results on 11 datasets and won 1st place on DocVQA (exceeding 90 for the 1st time) and WebSRC. Check Zero-shot Demo @huggingface
: http://
huggingface.co/spaces/PaddleP
addle/ERNIE-Layout
…