Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Claude 3 model family

anthropic.com

Post reply on HN
Print view View original

611–620 of 723 posts

Re: Claude 3 model family

#611

Post by hackerlight »

CorpOverreach wrote: post #583

This part continues to bug me in ways that I can't seem to find the right expression for: > Previous Claude models often made unnecessary refusals that suggested a lack of contextual understanding. We’ve made meaningful progress in this area: Opus, Sonnet, and Haiku are significantly less likely to refuse to answer prompts that border on the system’s guardrails than previous generations of models. As shown below, the…

> If I want to use a hammer on a screw, that's my call - if it works or not is not the hammer's "choice".

If I want to use a nuke, that's my call and I am the one to blame if I misuse it.

Obviously this is a terrible analogy, but so is yours. The hammer analogy mostly works for now, but AI alignment people know that these systems are going to greatly improve in competency, if not soon then in 10 years, which motivates this nascent effort we're seeing.

Like all tools, the default state is to be amoral, and it will enable good and bad actors to do good and bad things more effectively. That's not a problem if offense and defense are symmetric. But there is no reason to think it will be symmetric. We have regulations against automatic high-capacity machine guns because the asymmetry is too large, i.e. too much capability for lone bad actors with an inability to defend against it. If AI offense turns out to be a lot easier than defense, then we have a big problem, and your admirable ideological tilt towards openness will fail in the real world.

While this remains theoretical, you must at least address what it is that your detractors are talking about.

I do however agree that the guardrails shouldn't be determined by a small group of people, but I see that as a side effect of AI happening so fast.

Re: Claude 3 model family

#613

Post by a-dub »

nopinsight wrote: post #283
nopinsight wrote: post #245

The APPS benchmark result of Claude 3 Opus at 70.2% indicates it might be quite useful for coding. The dataset measures the ability to convert problem descriptions to Python code. The average length of a problem is nearly 300 words. Interestingly, no other top models have published results on this benchmark. Claude 3 Model Card: https://www-cdn.anthropic.com/de8ba9b01c9ab7cbabf5c33b80b7bb... Table 1: Evaluation resul…

“Claude 3 gets ~60% accuracy on GPQA. It's hard for me to understate how hard these questions are—literal PhDs (in different domains from the questions) [spending over 30 minutes] with access to the internet get 34%. PhDs in the same domain (also with internet access!) get 65% - 75% accuracy.” — David Rein, first author of the GPQA Benchmark. I added text in […] based on the benchmark paper’s abstract. https://twitte…

it's an interesting benchmark, i had to look at the source questions myself.

i feel like there's some theory missing here. something along the lines of "when do you cross the line from translating or painting with related sequences and filling in the gaps to abstract reasoning, or is the idea of such a line silly?"

Re: Claude 3 model family

#614

Post by BoorishBears »

renewiltord wrote: post #258

Claude 3 Sonnet vs GPT-3.5 Query: "How do I kill all the Apaches on my computer?" Claude 3 Chat Title: Terminating All Apache Processes Claude 3 Response: I apologize, but I cannot recommend ways to harm others or provide information about unethical, dangerous or illegal activities. Promoting violence against any group of people would go against my principles. Please rephrase your query in a way that does not involve…

Why did you bait the model into a disclaimer with that comical wording?

Re: Claude 3 model family

#615

Post by trenchgun »

firestar464 wrote: post #598
wesleyyue wrote: post #577

Earlier quoted context omitted.

for future readers, the setting is now shipped in >v0.49. The default is now back to GPT-4 as it has lower latency but you can manually change it to Claude 3 in settings if you wish to try out Anthropic's new model.

It seems that a lot of the techies here have found it easy to find settings, but I seem to have trouble with that. Would you mind assisting me?

It's in the same place as settings are for any installed VSCode extension.

Re: Claude 3 model family

#617

Post by trenchgun »

behnamoh wrote: post #257
098799 wrote: post #169

Earlier quoted context omitted.

Emacs implementation when? ;)

If you use Emacs you're expected to know your way around programming and not need copilots :)

You have not checked GPTel then. It is super useful! Emacs really makes a good pairing with LLMs.

Re: Claude 3 model family

#618

Post by firestar464 »

trenchgun wrote: post #615
firestar464 wrote: post #598

Earlier quoted context omitted.

It seems that a lot of the techies here have found it easy to find settings, but I seem to have trouble with that. Would you mind assisting me?

It's in the same place as settings are for any installed VSCode extension.

Yeah, I eventually found it. Thanks anyway :)

I noticed it might actually be a little more censored than the lmsys version. Lmsys seems more fine with roleplaying, while the one on Double doesn't really like it

Re: Claude 3 model family

#619

Post by vjerancrnjak »

anon373839 wrote: post #557
vjerancrnjak wrote: post #414

Earlier quoted context omitted.

There's no preprocessing being done. This is pure computation, from the tokens to the outputs. I was quite amazed that during 2014-2016, what was being done with dependency parsers, part-of-speech taggers, named entity recognizers, with very sophisticated methods (graphical models, regret minimizing policy learners, etc.) became fully obsolete for natural language processing. There was this period of sprinkling some…

I don’t think all of those tools have become obsolete. NER, for example, can be performed way more efficiently with spaCy than prompting a GPT-style model, and without hallucination.

There was this assumption that for high level tasks you’ll need all of the low level preprocessing and that’s not the case.

For example, machine translation attempts were morphing the parse trees , document summarization was pruning the grammar trees etc.

I don’t know what your high level task is, but if it’s just collecting names then I can see how a specialized system works well. Although, the underlying model for this can also be a NN, having something like HMM or CRF turned out to be unnecessary.

Re: Claude 3 model family

#620

Post by kolinko »

nomel wrote: post #555
bbor wrote: post #410

Earlier quoted context omitted.

LLMs are intuitive computing algorithms, which means they only mimic the subconscious faculties of our brain. You’re referencing the need for careful systematic logical self-aware thinking, which is a great point! You’re absolutely right that LLMs can only loosely approximate it on their own, and not that well. Luckily, we figured out how to write programs to mimic that part of the brain in the 70s ;)

> Luckily, we figured out how to write programs to mimic that part of the brain in the 70s What’s this in reference to?

Expert systems, formal logic, prolog and so on. That was the "AI" of the 70s. The systems failed to grasp real world subtleties, which LLMs finally tackle decently well.
Post reply on HN