armchair engineer. non-friendly neighbourhood tech support. annoying tweets & non-sense threads.
Joined April 2019
- Tweets34.1K
- Following4.3K
- Followers409
- Likes239K
Akhirnya. Berhasil dapet bahan buat nge-blackmail.
Ada 2 mingguan nyari-nyari info.
Terus begadang dari hari sabtu kemaren sampe hari ini tadi, tiap jam 23.00 - 05.00 buat ngikutin orangnya. Mana anginnya dingin banget anjg 💀
Passive income 5 jt per bulan bisa ini 😭😭
RaBitQ works by applying a random orthogonal transformation to a base scalar codebook & consequently multiplying the input vector by the inverse of that matrix to find the nearest quantised coordinate
Basically, turboquant works by applying a random rotation to the input data to force its coordinates into a predictable statistical distribution → independent 1D scalar quantisation
The TurboQuant paper (ICLR 2026) contains serious issues in how it describes RaBitQ, including incorrect technical claims and misleading theory/experiment comparisons.
We flagged these issues to the authors before submission. They acknowledged them, but chose not to fix them. The paper was later accepted and widely promoted by Google, reaching tens of millions of views.
We’re speaking up now because once a misleading narrative spreads, it becomes much harder to correct. We’ve written a public comment on openreview (openreview.net/forum?id=tO3A…).
We would greatly appreciate your attention and help in sharing it.
🧵 The Claude Sonnet 4.6 Thinking has FAILED to fix a simple AVC-to-H265 (NVENC) transcoding error.
1) Dia GAK nyadar kalo problemnya cuma di jpeg thumbnail yang di dalem container mp4 💀
2) nambah²in subprocess call ffprobe buat ngecek embd subtitle. buat apaan anjg??? 😭😭
> be veritasium
> no derek
> consists of 482 producers
> no derek
> made a video about asbestos
> didn't say the word "asbestos" for the first 9 minutes.
NINE WHOLE FUCKING MINUTES of vaguexplaining lmao 😭😭😭
youtu.be/cMx139eTxoc?si=GtfE…
✅ they acknowledge the Standard Model is already correct.
basically, they made it clear that they are trying to to understand the mathematical efficiency and structure of the theory instead of trying to fix an error in experimental prediction.
I genuinely COULDN'T find any flaw in the paper.
✅ the authors limit the scope to "Klein Space" immediately, in the abstract.
GPT-5.2 derived a new result in theoretical physics.
We’re releasing the result in a preprint with researchers from @the_IAS, @VanderbiltU, @Cambridge_Uni, and @Harvard. It shows that a gluon interaction many physicists expected would not occur can arise under specific conditions.
openai.com/index/new-result-…
