amilios

14 ms·

61.
by amilios 3y ago
OK, thanks for letting me know! It's a bit tricky to upload ArXiv links, I have a sense that not a lot of people will actually click through to read the paper or that some people just won't bother to click an ArXiv link at all, so
62.
by amilios 3y ago
Paper abstract: We attribute grokking, the phenomenon where generalization is much delayed after memorization, to compression. To do so, we define linear mapping number (LMN) to measure network complexity, which is a generalized version of
63.
Grokking as Compression: A Nonlinear Complexity Perspective(arxiv.org)
2 points by amilios 3y ago|1 comments
64.
by amilios 3y ago
At this point I doubt any language will be able to oust Python as the go-to language for machine learning. The Python ecosystem is just too strong, the value-add for the new language would have to be extremely high, I just don't think
65.
by amilios 3y ago
Ah I see. Yeah the criticism seems fair, gotta at least prepend "Abstract:" or something haha.
66.
by amilios 3y ago
The abstract of a paper without quotations? No, I don't think so. What do you mean exactly? Someone posted the abstract as a text post directly?
67.
by amilios 3y ago
Paper abstract: Large Language Models (LLMs), renowned for their remarkable performance across diverse domains, present a challenge due to their colossal model size when it comes to practical deployment. In response to this challenge, effor
68.
Outlier Weighed Layerwise Sparsity: A Missing Secret Sauce for Pruning LLMs(arxiv.org)
5 points by amilios 3y ago|1 comments
69.
by amilios 3y ago
Can anyone corroborate this anecdotally? I.e. has anyone actually looked at the output of the two models side-by-side for common tasks? There's lots of talks these days about academic benchmarks being pretty "broken" for mode
70.
by amilios 3y ago
Paper abstract: Recent advances in Language Model (LM) agents and tool use, exemplified by applications like ChatGPT Plugins, enable a rich set of capabilities but also amplify potential risks - such as leaking private data or causing finan
71.
ToolEmu: Identifying the Risks of LM Agents with an LM-Emulated Sandbox(github.com)
4 points by amilios 3y ago|2 comments
72.
by amilios 3y ago
Abstract: Large Language Models (LLMs) have achieved remarkable success in reasoning tasks with the development of prompting methods. However, existing prompting approaches cannot reuse insights of solving similar problems and suffer from a
73.
Thought Propagation: An analogical approach to complex reasoning with LLMs(paperswithcode.com)
56 points by amilios 3y ago|13 comments
74.
I think about LLM prompt engineering(fchollet.substack.com)
2 points by amilios 3y ago|0 comments
75.
The Dawn of LMMs: Preliminary Explorations with GPT-4V(ision)(arxiv.org)
2 points by amilios 3y ago|0 comments
76.
Agent Instructs Large Language Models to Be General Zero-Shot Reasoners(arxiv.org)
5 points by amilios 3y ago|0 comments
77.
Low-Resource Languages Jailbreak GPT-4(paperswithcode.com)
3 points by amilios 3y ago|0 comments
78.
Tora: A Tool-Integrated Reasoning Agent(github.com)
10 points by amilios 3y ago|0 comments
79.
by amilios 3y ago
Suffice to say I think this would only work in places like Europe where everything is relatively densely connected by rail. For me personally I live in North American and I visit my family in Europe once a year. Cutting me off from them wou
80.
by amilios 3y ago
...good? Isn't this kind of the point of Five Eyes? Seems like it's working as intended. In this case Canada is extremely justified to use whatever intelligence they have at their disposal I think, this was an extrajudicial killin
81.
by amilios 3y ago
Are people still using C# outside of like, corporate settings? I have the impression that especially with the recent Unity scare market share for C# must be trending downwards...
82.
41% of French pop in favour of limiting everyone to 4 flights for entire life(twitter.com)
33 points by amilios 3y ago|82 comments
83.
by amilios 3y ago
Not sure if this counts, but many successful games have been made with Game-maker-style tools, which I think would be classified as "low-code". E.g. Undertale
84.
by amilios 3y ago
Oh yeah that's true! Just block the stop token. But yes, my thought is that there are scenarios where it can get "stuck" as you said. I'll look at the PR, thanks!
85.
Emu: Enhancing Image Generation Models Using Photogenic Needles in a Haystack(twitter.com)
1 points by amilios 3y ago|0 comments
86.
Mistral 7B Instruct Unsafe(twitter.com)
2 points by amilios 3y ago|0 comments
87.
by amilios 3y ago
How does this actually work though, since the model could e.g. abruptly end the generation giving something syntactically invalid? Doesn't it need to look at the whole output at some stage?
88.
by amilios 3y ago
Huh! Nevermind then! I take it back. Would be interesting to see what kind of tuning they did/pit the model head-to-head with LLaMA-2-7B-chat. Seems like they did just instruction tuning but not RLHF? So I assume Mistral won't be
89.
by amilios 3y ago
Not a fair comparison as Mistral is not instruction tuned. You need to write a "continuation-style" prompt or few-shot it, rather than just giving it a direct order. Honestly this is one of the reasons imo that Mistral won't
90.
by amilios 3y ago
The technical reason it can't be done (or would be very difficult to do) is that weights are typically very uninterpretable. There aren't specific clusters of neurons that map to one concept or another, everything kind of does eve

More ›