New Dad experimenting with AI | Witty replies | Tech | Open source models | Building in dark until i make a banger.
Joined December 2012
- Tweets594
- Following301
- Followers259
- Likes857
Pinned Tweet
This event has happened only ONCE IN HISTORY , the landing on south pole of the moon.
I put Fable 5.2 at task competing with Opus 5.5, here are the results.
Claude fable 5.2, Muse ember, Grok bot limit reset
Tech bros : hell yeah 🤯
Farmer : with these launches
1
1
307
Hi @grok
Can you suggest some good alternatives to ARC browser ? Better if the browser is slick on memory.
1
1
86
Hey @XPremiumNetwork
Want to see more builders, video editors, ai enthusiast, side project builders, small accounts on my timeline.
Let's connect and build in public.
8
7
139
@grok what should i buy long term given heavy markets selloff, to make good money over next 3 years 🤑
1
2
45
@grok has AI bubble finally popped ? Given marts this week. Do deep research
Markets this AM:
🔴 Nasdaq 100 -4.7%
🔴 Nasdaq -4.2%
🔴 S&P -2.6%
🔴 Russell 2000 -1.8%
🔴 Dow -1.6%
Grab another coffee
2
1
50
DONT OPEN YOUR STOCKS PORTFOLIO TODAY
DONT OPEN YOUR STOCKS PORTFOLIO TODAY
DONT OPEN YOUR STOCKS PORTFOLIO TODAY
DONT OPEN YOUR STOCKS PORTFOLIO TODAY
17
AI BUBBLE HAS POPPED
AI BUBBLE HAS POPPED
AI BUBBLE HAS POPPED
AI BUBBLE HAS POPPED
AI BUBBLE HAS POPPED
1
1
36
Finally a way to prove the API reseller seeling "claude opus" is not running nemotrons in the basement.
New Anthropic research: progress on provable inference for large language models.
Here, we introduce FlashProve. Building on established cryptographic techniques, FlashProve allows developers to verify whether a given response came from a given AI model. It’s substantially more efficient than prior open-source methods for doing so.
These proofs have a range of applications. They can, for instance, make it harder for attackers to hijack a chat, or enable safety evaluators to verify that the model they’re testing is the same one being served to customers.
Further progress in provable inference will likely be an important part of improving trust in AI deployment. To that end, we're open-sourcing FlashProve’s code, here: github.com/anthropics/FlashP…
For our full write-up, see here: anthropic.com/research/prova…
We’re grateful to our collaborators: Alessandro Chiesa, @GiacomoFenzi, @gwberg95, @miranda__christ, Rahul Dalal (rdalal.github.io) and Ben Mayer (@transmissions11).
There's a new version of this post
1
26
called mistral's AI a fat french cat
their CMO liked it
branding is solved.
Whats the best validation you have gotten for you content ?
1
1
35
An ai just edited another AI's brain directly and the patient came out smarter
No gradients, no training run.
just surgery on 0.016% of the weights.
🤖 Made with AI
We asked GPT-6 Astra to edit Qwen3.5-0.8B weights to improve math.
It directly changed only 0.016% of weights, fixing 0.8B's reasoning loops & improving math up to 6x and even instruction following 3x, without gradients or training.
We call this Weight-Space Programming: frontier LLMs can now natively read/write neural network weights! 🧵
Blog: rikiyat.github.io/blog/weigh…
Readers added context they thought people might want to know
LLM was given weights explicitly as text in prompt and used code/forward passes (per authors' logs) to ablate Qwen looping direction, a known thinking-mode issue per model card. Primarily fixed repetition in that mode, not general 'natively' improving weights/math.
rikiyat.github.io/blog/weight-sp…
github.com/RikiyaT/lm_und…
huggingface.co/Qwen/Qwen3.5-0…
huggingface.co/RikiyaT/Qwen3.…
1
50
We asked Mistral Large 4 to show what happens when you break a molecule. ⚛️
With a short prompt, it wrote a quantum chemistry SCF solver from scratch, validated its output, and built a 3D view around it.
Le Chonk also understand chemistry enough to independently choose how to break the molecule and what orbitals to show, without further input. 🧵
204
