Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
guilamu
searching Neon…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
7 ms
·
1.
▲
Right Click XLSX <-> CSV PowerShell Script
(github.com)
2 points
by
guilamu
5d ago
|
1 comments
2.
▲
by
guilamu
5d ago
Extremely useful for me at work, thought I'd share: One 53 KB PowerShell file, 0 dependencies. Designed by Opus 5, two passes of bug fixes with GPT 5.6 Terra and Kimi k3 at max reasoning effort.
3.
▲
by
guilamu
23d ago
Related : https://m.youtube.com/watch?v=Q9uefFYe6bM&pp=0gcJCXACo7VqN5t...
4.
▲
by
guilamu
1mo ago
For some reason this page exists only in French, but it says a lot about the credibility of the ADL: https://fr.wikipedia.org/wiki/Anti-Defamation_League#Controv...
5.
▲
YouTube Mistakenly Penalizes Kurzgesagt for AI-Generated Slop
(kotaku.com)
29 points
by
guilamu
1mo ago
|
4 comments
6.
▲
by
guilamu
1mo ago
Where does this "10x" comes from?
7.
▲
by
guilamu
2mo ago
FTFY: Starting July 20 Claude Fable 5 will ve excluded from Pro plans.
8.
▲
by
guilamu
2mo ago
https://phosh.mobi/faq/#so-what-phones-are-supported
9.
▲
by
guilamu
3mo ago
Wouldn't you say that Valve is an exception to that rule?
10.
▲
by
guilamu
3mo ago
Any source to backup this claim, pretty please?
11.
▲
by
guilamu
4mo ago
If you're concerned about that do not give internet to your tv and use any kind of tv box instead (shield tv, apple tv, etc).
12.
▲
by
guilamu
4mo ago
It's not. France: €0.149/kWh (~$0.175) US: ~$0.12–$0.14/kWh https://www.globalpetrolprices.com/France/electricity_prices...
13.
▲
by
guilamu
5mo ago
You're right, I've certainly been a bit presumptuous to call this'a benchmark'. It is indeed a flawed test. Yet,It's been giving me the occasion to try some open source models and for my workflow, some of them are i
14.
▲
by
guilamu
5mo ago
Most people, including me, beg to disagree. Better Call Saul was a masterpiece. https://www.metacritic.com/tv/better-call-saul/
15.
▲
by
guilamu
5mo ago
Yeah as I said this a benchmark for my usecase only, a single use case, which is obvisouly not representative of everybody's needs. What strike me as very strange though is that 0 model were able to just use the search input already pr
16.
▲
by
guilamu
5mo ago
https://openrouter.ai/openai/gpt-5.5-pro 30/180 usd on Openrouter. Did I miss something?
17.
▲
by
guilamu
5mo ago
When nothing is noted it's max reasoning (xhigh in copilot chat in vscode if available). The models not availble on copilot were tested through opencode (max reasoning) and deepseek v4 was tested through Cline (with max reasoning too).
18.
▲
by
guilamu
5mo ago
Yes those two models were tested on my own PC (local inference using my own CPU/GPU). So something my be bugged on my setup. gemma4-26b should be far better than gemma4-e4b.
19.
▲
by
guilamu
5mo ago
Yes, the prompt is slim by design. I might be wrong, but the point was to see what the model can do "on it's own". The eval prompt is quite extensive: https://github.com/guilamu/llms-wordpress-plugin-benc
20.
▲
by
guilamu
5mo ago
Haha, just fixed the date! I haven't evaluated the judge benchmark. You have everything needed in the repo to do so though, so be my guest. It took me a bit of time to put all this together and won't have much more time to dedicat
21.
▲
by
guilamu
5mo ago
Yes Opus 4.7 fast (no reasoning) did a worst job than Sonnet 4.6 high (with reasoning) according to Gemini 3.1 Pro evaluation.
22.
▲
by
guilamu
5mo ago
Just tested it on my homemade Wordpress+GravityForms benchmark and it's one of the worst model of the leaderboard performance wise and the worst value wise: https://github.com/guilamu/llms-wordpress-plugin-benchmar
23.
▲
by
guilamu
5mo ago
Good point! I think I won't change anything right now or I'll have to remake all tests... I'll use your input for the Level 2 task I plan on working on.
24.
▲
Show HN: I blind-tested 14 LLMs on a WP plugin task. Surprising Findings
(github.com)
3 points
by
guilamu
5mo ago
|
2 comments
25.
▲
by
guilamu
5mo ago
Indeed. I'm still sad thought :(
26.
▲
No more Opus for Copilot Pro plan users
(github.blog)
37 points
by
guilamu
5mo ago
|
5 comments
27.
▲
Reasons to think that the Claude Mythos announcement was overblown
(garymarcus.substack.com)
2 points
by
guilamu
5mo ago
|
0 comments
28.
▲
by
guilamu
6mo ago
"Proton Mail, one of the services he moved to, is ultimately controlled by the US Gov," Would you mind elaborating, pretty please?
29.
▲
by
guilamu
6mo ago
Why not just get the iso, install, activate with massgravel and be done for life?
30.
▲
by
guilamu
6mo ago
That's true indeed, but Microsoft is not giving us any other option so why not use the good version at home? I mean what is the risk really?
More ›