Exciting but overlooked that ChatGPT is primarily an alignment advance—the base model (GPT-3.5) has been available in publicly for many months, but making it into a useful chat system required significant strides with reliably following the intent of the developer and the user.
LLMS
-
Correction on input length limit and token truncation
By
–
My example using "A A A A" contains a mistake — 2048 tokens is actually longer than the per-cell maximum input length. The input is silently truncated when too long. You need to give it 7 or 8 repetitions of an input that contains ~1024 tokens.
-
GPT-JT: Open Source Alternative to GPT-3
By
–
Meet GPT-JT, the Closest Open Source Alternative to GPT-3 https://
analyticsindiamag.com/meet-gpt-jt-th
e-closest-open-source-alternative-to-gpt-3/
… #midjourneyAi #dalle2 #Web3 #digitalartists #100DaysOfCode #OpenSource #bot #python #Digital #javascript #CodeNewbie #NFTs #digitalart #NFTArts #NFTartists #art #Artists #AI #DL #RT #AIart #ML -
GPT-3 Evolution: Open Source Models Outperform Proprietary Solutions
By
–
Today’s gpt3 is not the same thing as what was released two years ago and there are many better oss models for a lot of different tasks/setups/usecases (which is why most companies today use open models)
-
How to continue after hitting per-cell output limit
By
–
See how that last screenshot ends in a comma? It hit the per-cell limit on output length. All you have to do is ask it to keep going:
-
Foundation Models Double Every 2 Months Outpacing Moore’s Law
By
–
Doubling every 2 years: Moore's Law
Doubling every 2 months: Foundation models https://
nature.com/articles/s4158
6-022-05340-6
…
From deep to dendrocentric learning @Nature by @boahen_k -
AI hallucinates similar Spanish word from made-up input
By
–
You asked it for it to repeat your word in Spanish, and then gave it a completely made-up word, so it guessed a similar-looking Spanish word. That’s ordinary hallucination stuff — out-of-distribution inputs do that.
-
Long input instructions followed, truncation between 1024-2048 tokens
By
–
This isn’t true. I have examples in this thread of 1024+ token inputs and instructions at the end of them are followed. Truncation appears to happen somewhere between [1024, 2048] — I suspect it’s 2048 but haven’t confirmed.
-
AGIChat: Human-like AI responses on blockchain
By
–
AGI Chat "AGIChat is capable of generating human-like responses to a wide range of topics" – ChatGPT AGIChat.Eth : https://
opensea.io/assets/ethereu
m/0x57f1887a8bf19b14fc0df6fd9b2acc9af147ea85/28255763265962602015666407802618350348681692708073377058314554773073936729390
… #AGI #AGIChat #AGIFirst -
Model learns to handle phrasing errors in a general sense
By
–
It’s clearly learned to do this in a more general sense, as it should — e.g., this question contains a clear error in phrasing, but no human would respond to it literally as written. The model is trained on human demonstrations, so it tries do what humans do.