That was my mission, Matt. We're all gonna level up on ChatGPT together.
LLMS
-
Improving AI Responses Through Better Prompts and Randomization
By
–
Gotta give it a little bit more to work off of, otherwise you'll just get the most average, expected results possible. I'm actually trying to figure out a way to randomize the responses more. Will keep you posted.
-
GPT capabilities vary by topic, expected improvements ahead
By
–
For some topics, it seems to have more knowledge. Others, you sometimes have a harder time getting it to act more uniquely. But I do agree, it should get better over time. If OpenAi's GPT-3 is any indication, it'll see huge upgrades multiple times a year.
-
AI Output Eloquence vs. Reasoning: The ‘Nonsense Conclusion’ Effect
By
–
Agreed — but it's also somewhat independent. E.g., think of ChatGPT outputs that eloquently reach a nonsense conclusion by confidently asserting things that don't follow. There's a "Hey, wait a minute…" effect when style exceeds reasoning.
-
Stanford Buffalo Universities Advance Language Modeling State Space Models
By
–
Stanford & Buffalo U Advance Language Modelling with State Space Models https://
syncedreview.com/2023/01/03/sta
nford-buffalo-u-advance-language-modelling-with-state-space-models/
… -
AI Dungeon’s Role in the Birth of LLM Prompt Engineering
By
–
LLM prompt engineering was born in academia, but AI Dungeon was its cradle. Many people, including me, used GPT-3 for the first time via this game. Many useful prompting techniques were discovered by players who abused it for independent, informal research.
-
AI Leaders Share Hopes and Key Priorities for 2023
By
–
1/Several leaders share their hopes for AI in 2023, including finding key missing pieces that will enable algorithms to reason, building a personal data timeline, improving AI processes, discovering new principles for explainability, and using generative AI for active learning.
-
Interest in Small Local Language Models Under 1 Billion Parameters
By
–
i’d LOVE a cheap small (sub 1B) model to run this stuff locally! actually am wondering if gpt2 does this well enough
-
ChatGPT discoveries from Reddit and Discord communities shared
By
–
I've been poking around in various Reddit communities and Discord servers, and oh man, there's some wild stuff going on with ChatGPT. Trying to bring the best of that back to the masses and make it more accessible.
-
ChatGPT’s Untapped Potential and Hidden Capabilities
By
–
So many interesting things to try. And many more not even listed here. It’s so complex, even the people who designed ChatGPT don’t fully know yet what it’s capable of