This is a feature that is available to Pro accounts where they can switch between Haiku and Opus. It is also available as a feature flag which I have modified but if I select the Opus model it won't work for me. There is also a debug button
LLMS
-
Comparison of AI model performance and usage limitations
By
–
Yeah, feels quite limited in comparison to ChatGPT, especially considering their new dynamic limits
-
Companies Adopting LLMs as Documentation for Software Tools
By
–
I am waiting for the first companies to seriously adopt LLMs-as-documentation. Especially for software tools, they can get you remarkably far. https://t.co/EgBmm3bQvZ
— Ethan Mollick (@emollick) 8 avril 2024I am waiting for the first companies to seriously adopt LLMs-as-documentation. Especially for software tools, they can get you remarkably far.
-
LLM Consciousness Claims: Evidential Value Compared to Dream Characters
By
–
A claim of consciousness from an LLM has no more evidential value than the same from a character in a dream. The latter is more plausible a priori as the hardware is known to support it.
-
LLMs Can Solve More Problems Than People Realize
By
–
It doesn’t mean that LLMs can solve every problem, by the way. But they can do a lot more than people think.
-
AI Struggles with Word Games Due to Tokenization Limitations
By
–
AI is very bad at word games. Tokens get in the way.
-
Non-Techies Self-Hosting LLMs: Reality Check
By
–
Everyone who thinks non-techies are going to self-host LLMs is sorely out of touch with reality
-
Developing new AI writing tools for TestingCatalog
By
–
My new set, working on more advanced AI writers for TestingCatalog
-
LLM Context Window and Program Execution Simulation Debate
By
–
In light of this comment, what is the point of the contest? An LLM with sufficient context window can simulate execution of any program — what does it matter whether some given program’s exec trace fits in 32K when no one would ever use a pure LLM for this vs. an LLM + REPL?
-
Analysis of Claude’s current competitive value and feature gap
By
–
There are no reasons to pay for Claude atm unless you want to use the model directly via other platforms. It will take Anthropic a long time to bridge a feature gap with other competitors (even ignoring this unnecessary ban wave).