tartakovsky

I am curious how this is different than me installing Hermes on a VPS. It seems I can do anything from there.

Is Claude self-replicating in an attempt for world domination?

Well, task == Resolving real GitHub Issues

Languages == Python only

Libraries (um looks like other LLM generated libraries -- I mean definitely not pure human: like Ragas, FastMCP, etc)

So seems like a highly skewed sample and who knows what can / can't be generalized. Does make for a compelling research paper though!

2-ish questions:

Is this level of fear typical or reasonable? If so, why doesn’t Anthropic / AI code gen providers offer this type of service? Hard to believe Anthropic is not secure in some sense — like what if Claude Code is already inside some container-like thing?

Is it actually true that Claude cannot bust out of the container?

This landing page also fails to properly describe anything I understand.

I clicked 3 different links failing at answering that question for myself and then I stopped caring, though not enough to pass up one-upping your comment.

original paper: https://news.ycombinator.com/item?id=41784591

Huh? “Total Shazams ever detected: 240. That's an average of 240 songs per day.”

spoiler alert -- no need to share the answer, it takes the fun away for others

ollama is MIT licensed unless i am misreading

What is your goal? if d1, d2, d3, etc is the dataset over which you're trying to optimize, then the goal is to find some best performing d_i. In this case, you're not evaluating. You're optimizing. Your acquisition function even says so: https://rentruewang.github.io/bocoel/research/

And in general if you have an LLM that performs really well on one d_i then who cares. The goal in LLM evaluation is to find a good performing LLM overall.

Finally, it feels that your Abstract and other snippets sound like an LLM wrote them.

Good luck.

What is this exactly in plain english, please?

A happy person is not in a particular set of circumstances, but rather has a particular set of attitudes.

The webpage discusses a Bayesian approach to experimentation, focusing on interpreting and extrapolating experimental results, mainly in a tech environment aiming to maximize user retention. It addresses challenges like the inference problem, the extrapolation problem, the explore-exploit problem, and a culture problem within tech companies around misuse of experiments. The author suggests providing decision-makers with benchmark statistics to help them estimate true effects of different policies, and discusses a model of experimentation dealing with observed and true effects along with the noise in experiments 1 .

Great question. A legend or brief description of the underlying logic / heuristic would be helpful.

Not secure... NET::ERR_CERT_COMMON_NAME_INVALID Subject: *.safezone.mcafee.com

Issuer: McAfee OV SSL CA 2

Expires on: Aug 3, 2023

Current date: Jul 8, 2023

PEM encoded chain: -----BEGIN CERTIFICATE----- MIIGfzCCBWegAwIBAgIQKt9VNrFtaozA1bILX1OcfzANBgkqhkiG9w0BAQsFADBk MQswCQYDVQQGEwJVUzELMAkGA1UECBMCQ0

What is “human intent”?

Why do you not think it’s featured? Has a16z funded many of those companies, lol? And somehow, rejected milvus?

Same idea here? Larger models do a better job forgetting their training data and dropping their semantic priors. Perhaps another way of thinking through this is that larger models learn new information and drop old information faster. https://arxiv.org/abs/2303.03846

Isn't that interesting? The idea of "mental liquidity", or "strong opinions weakly held"? https://news.ycombinator.com/item?id=36280772

Related: https://medium.com/@ameet/strong-opinions-weakly-held-a-fram...

What’s better, train a model with 10X parameters once on some default hyperparameter setting or to search for a good hyperparameter configuration by training on X parameters 10 times? While I’m at it, how many LLMs of the size of GPT3 were trained until they landed on the capability of GPT3? How much of this is dependent on the data, or do good settings transcend the type of text that a model is trying to train on?

Strange how at the top of that link it shows `Model: Default` but in the actual session it shows `Model: GPT-4`. An OpenAI display bug, I would assume: https://www.dropbox.com/s/f2k1c7kaijjmpzu/Screenshot%202023-...

Just having a simple conversation with ChatGPT (the upgraded GPT-4 version) this morning. In the end searching GitHub myself was faster. Any tips on this or research on this topic? Is this what is referred to as In Context learning? Or In Context reminders?

This reminds me of this article for which symbol fine tuning helps overcome: https://arxiv.org/abs/2303.03846

Basically they show that larger models have an easier time giving up their semantic priors, so that, in the context of OP paper, learning the mapping for len becomes print and print becomes len.

Is this different from https://langchainx.web.app/? How?

Made me laugh because it's true, funny. Well done!

I’m just saying, it was more a hopeful nod in the direction of wishful thinking. Not a joke, just a bit of having my head in the sand.

Wow! I don't know how accuracy translates, I do see charts that look strong but unless I'm missing something, this is incredible. Would be curious and also terrified to see an endpoint so I can play around with it. I thought we were stopping this kind of research?

Just when OP wrote: 'Griptape can be thought of as "Airflow for LLMs," providing an alternative to the agent-based LangChain approach.'

This looks cool. I am sort of unclear about this contrasting statement to Langchain when it seems you are building on too of / or using Langchain to achieve some functionality. Is that incorrect? Also, I looked at your repo to contribute, but sadly I see no issues!