iamskeole

5 ms·

1.
Show HN: Don't trust me bro: fixing GPT-OSS (3.49B tokens, 1k GPU hours, 1x3090)(github.com)
2 points by iamskeole 16d ago|0 comments
2.
OpenAI COO Resigns(xcancel.com)
12 points by iamskeole 1mo ago|1 comments
3.
by iamskeole 1mo ago
I'm surprised by the sample efficiency, I thought that's still a largely unsolved problem. From the video: > [14:00] Jun 11: Highly persistent experimental internal-only model begins training and during this time the SSRF vulne
4.
by iamskeole 2mo ago
here you go: https://huggingface.co/blog/agent-intrusion-technical-timeli...
5.
Genesis chip may help AI with its memory problem(techxplore.com)
1 points by iamskeole 2mo ago|0 comments
6.
Opus 5 ARC-AGI-3 likely benchmaxxed(xcancel.com)
4 points by iamskeole 2mo ago|2 comments
7.
Predicting AI Job Exposure(ben-evans.com)
2 points by iamskeole 4mo ago|0 comments
8.
by iamskeole 4mo ago
You can also download the same model and run it locally without data leaving your machine, the Zed tram released it as open weights: https://huggingface.co/zed-industries/zeta-2
9.
by iamskeole 5mo ago
HN seems to redirect to old.reddit, which doesn't seem to show images on my phone at least, and the author included an image [1] of the model selection in one of the subsequent comments. Anybody able to verify the claim? [1] https:&#x
10.
by iamskeole 5mo ago
My first reaction as well. There's a photo of the user's screen in the post as well, which doesn't look fabricated.
11.
Claude Code Uses GLM 4.7(old.reddit.com)
2 points by iamskeole 5mo ago|4 comments
12.
by iamskeole 6mo ago
Are there any plans for QAT / MXFP4 versions down the line?
13.
by iamskeole 6mo ago
It seems the vote passed [1], meaning the existing Regulation [(EU) 2021/1232] was _extended_ until August 3 2027, with some amendments to the previous text: - added targeted scanning requirement - scanning must be “targeted, specified
14.
by iamskeole 6mo ago
can you describe your pipeline(s)? eg.: how do you decide topics to write on? are you just a simple loop searching for "news today" then rewriting with a twist or something else? go wild, this is a forum where people embrace the t
15.
Frontier AI Models Exhibit Sophisticated Reasoning in Simulated Nuclear Crises(arxiv.org)
3 points by iamskeole 7mo ago|0 comments
16.
How will OpenAI compete?(ben-evans.com)
482 points by iamskeole 7mo ago|671 comments