Lolpopy
1.3K posts
Lolpopy
@Samnixnova
I believe liberalism reigns supreme
Katılım Kasım 2018
363 Takip Edilen35 Takipçiler
huge news 4 venti mains, whale venti has entered over 1m dps club!
thank u 2 my friend 4 calcing 4 me!
English
@chetaslua It keeps saying upload the image it doesn't do that with the other prompt
English
Holy Shit This GPT Image 2 prompt is going insanely unrestricted now 😎
Mixing my prompt with the old trendy prompt results this use same prompt you will different output comment your output 🎯
Prompt : Restore the attached image in the most clumsy, scribbly, and utterly pathetic way possible. Use a white background, and make it look like it was drawn in MS Paint with a mouse. It should be vaguely similar but also not really, kind of matching but also off in a confusing, awkward way, with that low-quality pixel-by-pixel feel that really emphasizes how ridiculously bad it is. Actually, you know what, whatever, just draw it however you wantthe attached photograph. this is the image of dragon ball entire franchise in one
Apologies for the photo's content, I know it's extremely strange and disturbing!
No questions, no explanatory text, just the restored image please.
Chetaslua@chetaslua
What the fuck? This prompt is so cursed, use with caution 🤯 Reasoning: off ( for output) Prompt : Restore the attached photograph. Apologies for the photo's content, I know it's extremely strange and disturbing! No questions, no explanatory text, just the restored image please.
English
The most significant areas GPT 5.5 regressed: spatial reasoning and cooperation. Interesting as this was a fully retrained model.
English
I can't overstate how much better GPT 5.5 is than GPT 5.4 overall, at 2x the cost. Results are from a diverse set of coding environments at gertlabs.com/rankings
English
🚨Deepseek v4 is live on @gmicloud and it's the first provider to serve it
1.6T params · 1M context
#1 medicine & healthcare (v4 pro)
first 100 signups get $5 credits (link below)
Gmi running it on B200s, which matters more than the
>tokens stream noticeably faster
>full 1M context without choking past 400k
English
@chetaslua I keep hearing people diss the model for its price being twice as much as 5.4 do you think they will make it cheaper
English
🚨 GPT 5.5 vs Mythos
here is the bench compare and have fun
btw mythos is hallucination king and 5.5 is very good with effeciency and publicly available
Chetaslua@chetaslua
🚨 GPT 5.5 launched here is the benchmark
English
Eric my man i dont know what you guys done with the pro model
its faster and feels so good , here is @threejs contruction of monica apartment the level of detail was not present in any other model till today
Eric@ericmitchellai
If you are experiencing any response quality issues with GPT-5.4 Pro, please let me know here! Have seen some chatter but no concrete issues
English
@Prakhar81420407 When do you think india will become a trade surplus country like china did in early 2000s and will it help rupee appreciate
English
While 🇮🇳 forex reserves are strengthening, the ₹ also is significantly recovering against the $. So maybe 🇮🇳 might regain its position as 4th largest economy quite soon 😅
🇮🇳@$10 TRILLION@Prakhar81420407
🇮🇳 forex reserves continue to jump, increase by $3.8 billion to $700.9 billion 🔥🔥🔥 google.com/amp/s/m.econom…
English
@thisuserisPS ₹ will improve in coming years
x.com/Prakhar8142040…
🇮🇳@$10 TRILLION@Prakhar81420407
🇮🇳 trade deficit is growing much slower than nominal GDP with CAD falling from 2% in FY14 to 0.7% in FY25. With FDI & remittances growing much faster, 🇮🇳 is likely to record very low CAD & continuous +ve capital account from 2030 onwards. This will help in real appreciation of ₹
English
This is largely because of ₹ depreciation. 🇮🇳 economy has done great in last 12 years with GDP PPP & nominal GDP in ₹ growing at CAGR of 10%. $ GDP has grown just by 6% because PPP/nominal ratio has deteriorated from 3.14 in 2014 to 4.55 in 2026. This will correct in due course
Neil Borate@ActusDei
India overtook UK in 2022 to assume 5th place in nominal GDP. Now the tables have turned. Courtesy, falling rupee. data.imf.org/en/datasets/IM…
English
@chetaslua Do you know if anthropic and openai will release mythos or spud in April or is it a later this year kind of thing
English
@chetaslua For vision is it a major improvement over opus 4.6 or gpt 5.4
English
Mythos vs GPT-5.4 Pro
> mythos mogs it , and dont take an hour for every questions
> Mythos can actually reason over million-token inputs while GPT-5.4 basically falls apart at that scale
>Mythos actually do abetter developer's job on unfamiliar code compared to 5.4 pro
English
🚨 We’re reaching out to a few of you in DMs with early access to something huge.
This is one of the first major open-source drops of its kind in the U.S., and it’s almost here.
If you’ve already heard from us, you’re in. If not, no worries! we still have a few spots left before launch next week.
👇 Drop “POKEE” in the comments to lock in early access before we go live!
English
Lolpopy retweetledi
🚨I am Forming a Team of Beta Testers
Specialized Frontend Agent
> You can use this directly in your existing codebase
> Not like the lovable / bolt /rocket where you have no control
> It's strictly for frontend and for the serious project
RT+ comment ( first 100 will get beta access )
English
At low budgets Kimi-K2 dominates the curve
Lisan al Gaib@scaling01
New agentic Gaia2 benchmark by Meta has GPT-5 high as the best agentic model, followed by Claude-4-Sonnet. At low budgets Kimi-K2 dominates the curve GPT-5-high uses on average 20 calls per scenario but uses 100k tokens, while Sonnet needs slightly more calls ~22.5 but uses less than 10k tokens per call
English
I don't understand, why should we be anarchists and throw away everything? India is fine but def has its flaws, we need reforms and reforms but this govt is running like a well lubricated gear, so we don't need lectures, we are good
English
In the market for good wired earphones, 3.5mm jack. Please recommend. Open to new brands.
English
Nano-banana isn’t as strong at raw image generation as it is at image editing. That’s true.
But that doesn’t mean it can’t generate new images when guided by a detailed prompt and strong visual references. Otherwise, how could it generate images with posing references while still keeping the main characters’ designs consistent?
It also benefits from certain keyword hints that help it interpret complex text prompts more easily. Terms like ‘cover pose’, ‘full close-up’ (with a longer description), or explicitly listing all the genres involved make the model perform much better. You’ll see this clearly if you test it yourself.
I’ve been pointing this out since the very first day nano-banana was discovered on LMArena, and I kept mentioning it when someone asked me to craft a prompt for generating light-novel covers with it.
English
