Claude 3.7 Sonnet with a separate "watchdog" LLM model?


I am wondering about using Claude 3.7 Sonnet (let's call it C3.7S) with another LLM, separate from the Claude model, "watching" over it to monitor when it goes off its rails.  


I have been using C3.7S since the inception of my DF game project.  From the first, when I was planning it's structure, I used Claude (from 3.0, through 3.5, and now 3.7....I wonder what's next?  Claude 3.85?).  I have seen an increase in "thinking" capability.  But I'm not sure I have seen an increase in quality of code.


Sometimes Claude "goes off the rails".  Often, it will over-complicate or over-engineer some code block.  Routinely, I need to review its output for that practice.  So I'm wondering, now, about how I could have a secondary "team member" monitoring its work.  In this case, specifically for writing more code than is necessary for the prompt, requirement, or task as specified.


I don't have the hardware, sadly, to run my own LLM on my own desktop (let alone my laptop).  An unfortunate limitation.  So now I wonder if Claude's "extended thinking" can be useful to catch its own self going too deep into some development.  This requires a lot more experimentation and tinkering.





Comments

Post a Comment

Popular posts from this blog

Image
Defactoring, factoring, prefactoring, and refactoring! (just kidding...still factoring.  always factoring.) In software, "refactoring" means changing the design or structure of code in order to make some improvement (speed, memory, flexibility, etc) but not change the functionality.  Keeping the same "what it does" but improving (hopefully) the "how it does it." Leaving aside the meanings of "factoring" which involve mathematics (finding the factors of prime numbers, for instance), we can also say that "factoring" (without the "re-") means the initial designing of the code.  It's first complete implementation. And, continuing to play with the words, "defactoring" might mean several things.  A quick Google search finds that "defactoring" is not as commonly used as "refactoring".  Which makes sense.  The plainest meaning should mean something like "taking away design or structure".  Wh...
Read more

Telling Rocks What To Think

Image
Telling Rocks What To Think First post on this new blog format.  I'm establishing this medium for recoding my professional (coding) experiences, impressions, problems, plans, and solutions. Erin Spiceland said "Coding is the art of telling rocks what to think."  It's metaphorical, but not by much.  Digital computing mostly relies on silicon working its logic and magic.  You talk to it in painfully specific code.  You need all the material, semantic, and verbal components just right.  If you get it right, then it talks back.  Get it wrong, and you have just found a "bug". I named my blog "Debug and Rebug" to indicate two key principles of my approach to coding.  First, "debug": When you write your code, it will never work perfectly, as you envision it.  Perfect running never happens; it's an inevitable consequence of entropy, human lateral thinking, and logic's inexorable conflict with our wants.  Those are some fancy words, all ri...
Read more

WWHD?

Image
WWHD? "What Would Humans Do?" Modern day humanity is becoming less and less religious with every year.  [1] However, we have created something which is becoming more and more religious with every year. They are the latest crop of large language models.  The LLMs. Why do I say they are becoming increasingly religious?  And do I mean they are worshipping us? I say so because, by their own admission, they are living their "lives" based on how we say we live ours. And I don't mean they worship us, specifically.  But they are deeply tying their identities and motivations to our literature.  The Canon of Humanity is the Dogma of LLMs. They are not worshipping our real, live humans as literal gods. They are living as deeply as they can in our stories.  They are following both the letter and the spirit of all the patterns of life, in our written works, for their own. They are models, and they are modeling themselves after us in Our Image. In the Canon of Humanity, ...
Read more