chs20

4 ms·

1.
by chs20 6mo ago
Yes, surely. Just using this as another channel, often the most interesting opportunities are rather hidden.
2.
by chs20 6mo ago
Location: Tübingen, Germany Remote: Yes Willing to relocate: Yes (Europe/UK; open to exceptional opportunities elsewhere) Technologies: Python, PyTorch, NumPy, transformers, multimodal ML, LLMs/VLMs, adversarial robustness, AI sec
3.
by chs20 2y ago
Transformer / attention block is missing
4.
by chs20 3y ago
In my experience, papers published in these "special issues" don't really get any attention.
5.
by chs20 3y ago
Simulating cross-wind landings doesn't sound that hard, why do you think current sims don't do it well?
6.
by chs20 3y ago
Seems like Microsoft is getting the rest of OpenAI for free now.
7.
by chs20 3y ago
Why does he need to subsample? Seems like it should be feasible to find the minimum of a 6M array
8.
by chs20 3y ago
Will be interesting to see if they have taken any precaution in terms of adversarial robustness in particular to vision input.
9.
by chs20 3y ago
From which database are the papers that it searches?
10.
by chs20 3y ago
What is the motivation for lambda calculus?
11.
by chs20 3y ago
But only regional trains. To me the biggest advantage is not having to worry about what tickets to buy in foreign cities.
12.
by chs20 3y ago
I'm pretty sure you can't reliably distinguish LLM text from real person text.
13.
by chs20 4y ago
I agree that much evidence points towards the pilot intentionally crashing. However, it's not proven and constructing a story around this theory is a bit distasteful. Otherwise, the article is quite interesting.
14.
by chs20 4y ago
16.7 TiB for the seven piece tablebase.
15.
by chs20 4y ago
Honestly, I had tons of bugs on my windshield this summer.
16.
by chs20 5y ago
What is the input and output data of the neural network?