HackerTrans
トップ新着トレンドコメント過去質問紹介求人

cjrd

no profile record

投稿

The "Super Weight:" How Even a Single Parameter Can Determine a LLM's Behavior

machinelearning.apple.com
2 ポイント·投稿者 cjrd·11 か月前·0 コメント

Changing a single number among billions can destroy an AI model

newscientist.com
3 ポイント·投稿者 cjrd·2 年前·1 コメント

Pruning a single weight in an LLM can destroy its output

arxiv.org
1 ポイント·投稿者 cjrd·2 年前·0 コメント

Accelerating Ukraine Intelligence Analysis with Computer Vision on SAR Imagery

bairblog.github.io
6 ポイント·投稿者 cjrd·4 年前·0 コメント

コメント

cjrd
·4 か月前·議論
now do one for human-coded incidents.
cjrd
·9 か月前·議論
Eh
cjrd
·昨年·議論
some benchmarks here: https://machinelearning.apple.com/research/apple-foundation-...
cjrd
·2 年前·議論
Really enjoyable overall.

As a noob, I did feel like I spent the majority of my time just running around trying to figure out what to do.
cjrd
·2 年前·議論
It's great to see a _real_ AI application among all this media noise ;-).

Seriously though, this is wonderful satire. I asked 88x10 and it returned an HTML meta tag.
cjrd
·2 年前·議論
That's the big problem? A bug made it into prod at ChatGPT and Google didn't have a good set of safety rails (does anyone, yet?)?
cjrd
·3 年前·議論
Let's check out the paper for actual tech details!

> Given both the competitive landscape and the safety implications of large-scale models like GPT-4, this report contains no further details about the architecture (including model size), hardware, training compute, dataset construction, training method, or similar.

- OpenAI
cjrd
·3 年前·議論
> Given both the competitive landscape and the safety implications of large-scale models like GPT-4, this report contains no further details about the architecture (including model size), hardware, training compute, dataset construction, training method, or similar.

Thanks OpenAI
cjrd
·3 年前·議論
This is all predicated on existing conditions, where AI-written text hasn't influenced the way that humans write. As the years pass and these tools become a common way to at least "spot check" your own writing, I imagine that we will all begin to write in styles that are increasingly similar to AI-written text.
cjrd
·4 年前·議論
First, is productivity really the issue here? It makes for a great sound bite, but I imagine we've all spent a lot of time and effort working really, really hard...on the wrong thing.

Second, for large companies that want to weather the "impending recession," how is it that working harder will allow them to do this? What specific results will this yield? More product launches/improvements? Happier customers because of these launches (heh - when was the last time this happened for these companies) that translates into more revenue?

What I would love to see are execs that say something like "We really want to focus on listening more to our customers and improving our relationship with them. While others are shouting 'build! build! build!', we're saying 'listen, build, repeat.' Here's some specific ways we are going to do this: ..."

Then, sure, turn up the heat internally around this mission. Great - a rally cry around an objective. But right now, the rally cry is the rally cry is the rally cry. Work hard to work harder so that we work harder, and oh yeah, we'll fire people who don't because they're lazy and not 1337 enough to be here. You know, because recession.
cjrd
·4 年前·議論
Ah yeah - I've been bummed that people now think https://metacademy.org/ is about meta.
cjrd
·5 年前·議論
A fox eats a rabbit out of necessity. In our modern economy, people don't need to eat animals; it's usually a matter of convenience, taste, and tradition. But in all reality, we can get by quite comfortably without doing so.