I haven't evaluated it yet, but I love the fact that the output is (at least claimed to be) deterministic. I can't trust an LLM to do the right thing after I deploy it to production, because their output is non-deterministic by design.
TERMy (or is it the NPC-forge) seems to be worth a try.
You can get determinostic output (mostly) by setting the temperature to zero. Using couple of other tricks you can get close to 100% of determinism with LLMs.
That's reproducible, I wouldn't call it deterministic. Small, semantically meaningless changes in the input can still result in wildly different output.
That's the definition of a chaotic system (small change in initial conditions results in large, seemingly -- but not actually -- random changes in output), but it's still deterministic (same input results in same output).
> because their output is non-deterministic by design.
It isn't. At least not by design, even though in practice it often can be. If you do greedy decoding (or use a preset seed) and deterministically compute everything (e.g. only use integer math) then it will be 100% always deterministic.
That’s true, but not true-true. Sure, every time you prompt “what is the weather in kansas” you’ll get the same output, but if you prompt “what is the weather in kansas right now” you’ll get a different output, and then “what is the weather in kansas today” gets a different output. Language being language, there are infinite ways to say things, so there are infinite variations in what the llm can output in response to very similar prompts.
This tool has a finite amount of outputs for an infinite amount of inputs. Which is different from an llm based tool.
Yes i think I acknowledged that, but is that useful for making a tool that can be trusted to safely run shell commands when asked arbitrary questions? No. It’s not.
There's a lot of people who know Excel and SQL who otherwise don't know anything else technical. By my measure what they do is programming but a lot of people don't see it that way.
Consent-O-Matic is a bit of an edge case because their last update was more than a year ago and their Github repo looks pretty dead. This makes me not really trust it anymore.
Interesting they'd recommend Decentraleyes: its default configuration disabled all HTTP link prefetching, which hurts performance in a way not apparent from the extension description.
Decentraleyes has worse issues than that, it’s been mostly abandoned for 7 years and never covered much [1]. LocalCDN is better, but coverage is still insufficient, it still has no chance of actually preventing a CDN connection more than once in a blue moon. Modern privacy folks don’t recommend tools like this because they don’t really work and they do make your network and timing signature much more unique and fingerprintable.
I seem to remember TamperMonkey being on there but not ViolentMonkey, which I didn't agree with. However, I see neither of them on there. Both are available but not recommended. I wonder what is going on with that.
I see GP's point tho. The original Greacemonkey was open-sourced. Tampermonkey was the successor and is proprietary. Violentmonkey came to fix that. I wouldn't be surprised if Mozilla rejected TM because of not being OS and rejected VM because of lack of popularity
I get what you mean, and it'd be cool if "Recommended" was only for FOSS, but I don't think that ever been the case? Probably lots of the "Recommended" extensions are proprietary, from https://addons.mozilla.org/en-US/firefox/search/?promoted=re... you don't get far before hitting one, the 3rd extension already is proprietary out of those, surely more.
I suspect they no longer recommend either because either can download/execute scripts that Mozilla can't review, and the recommended extensions are reviewed, so they don't want to give users the wrong impression.
TERMy (or is it the NPC-forge) seems to be worth a try.
reply