Revanth Atmakuri

6.9K posts

Revanth Atmakuri banner
Revanth Atmakuri

Revanth Atmakuri

@RevanthAtmakuri

Follow for more latest tech research

Ontario, CA Katılım Mayıs 2021
466 Takip Edilen144 Takipçiler
Revanth Atmakuri retweetledi
CG
CG@cgtwts·
GitLab founder Sid Sijbrandij was diagnosed with rare cancer called osteosarcoma. when doctors ran out of options, he used AI to research treatments, built his own medical team, and created a personalized therapy plan. people are now using AI to cure cancer this is INSANE.
orph@orphcorp

this is excellent >GitLab founder diagnosed with rare cancer (osteosarcoma) >standard care works but cancer comes back later >medical team says there's not much else to do >"It became my own job to keep myself alive. Nobody else was going to do it for me at this point" >starts researching, assembles his own medical team, uses AI for deep research >“I’ll talk to anyone, I’ll go anywhere, and I can be there anytime" to collect information >does as many diagnostic tests as he can find as often as he can (maximal diagnostics) >develops his own therapeutic ladder with repurposed drugs, personalized medicine, etc >Sid’s cancer currently in remission

English
22
71
939
141K
Revanth Atmakuri retweetledi
Min Choi
Min Choi@minchoi·
The future is here. Google Translate just turned your headphones into a personal translator in 70+ languages. Both Android & iOS is live now👇
Google@Google

Your headphones just became a personal translator in 70+ languages. 🎧✨ Google Translate’s “Live translate” with headphones is officially on iOS. We're also expanding this capability to more countries around the world for both @Android and iOS users. To try it, open the Translate app, tap “Live translate” and connect your headphones.

English
15
18
99
23.3K
Revanth Atmakuri retweetledi
Eyisha Zyer
Eyisha Zyer@eyishazyer·
Found this amazing site where you can build electronic circuits right in your browser and test them on the spot...
English
48
520
4.1K
187.5K
Revanth Atmakuri retweetledi
jenn ☀️
jenn ☀️@jennsun·
overheard a new insult: you have a short context window 💀
English
231
1K
12K
396.3K
Revanth Atmakuri retweetledi
Rohan Paul
Rohan Paul@rohanpaul_ai·
Another beautiful story, of AI diagnosis where doctors failed. Going wildly viral on Reddit. His 62-year-old uncle in India had dialysis, diabetes, hypertension, stroke history, and severe migraines that happened only when he lay down to sleep. Multiple specialists saw him. He had brain imaging and other treatment. But nobody could explain why the headaches were clearly positional. "Claude didn't just identify the problem. It created a structured diagnostic roadmap, explained which specialist to see first, what tests to request, what questions to ask, picked the right CPAP machine, explained every setting, and even wrote maintenance instructions in Gujarati (my native language). A $317 CPAP machine (advised by Claude) solved what years of specialist visits couldn't." --- reddit .com/r/ClaudeAI/comments/1s41fny/25_years_multiple_specialists_zero_answers_one/
Rohan Paul tweet media
English
51
105
873
93.3K
Revanth Atmakuri retweetledi
Nathie
Nathie@NathieVR·
Mixed Reality is no longer a gimmick
English
79
274
3.4K
400.7K
Revanth Atmakuri retweetledi
AI at Meta
AI at Meta@AIatMeta·
Today we're introducing TRIBE v2 (Trimodal Brain Encoder), a foundation model trained to predict how the human brain responds to almost any sight or sound. Building on our Algonauts 2025 award-winning architecture, TRIBE v2 draws on 500+ hours of fMRI recordings from 700+ people to create a digital twin of neural activity and enable zero-shot predictions for new subjects, languages, and tasks. Try the demo and learn more here: go.meta.me/tribe2
English
708
2.2K
14.3K
6M
Revanth Atmakuri retweetledi
Chubby♨️
Chubby♨️@kimmonismus·
Meta is back: Meta just dropped TRIBE v2, a foundation model that predicts how your brain responds to sight, sound, and language. Trained on 500+ hours of fMRI data from 700+ people, it can predict a new person's brain activity without any retraining, and its predictions are actually more accurate than a real brain scan. Neuroscience just got a serious AI upgrade!
Chubby♨️ tweet media
AI at Meta@AIatMeta

Today we're introducing TRIBE v2 (Trimodal Brain Encoder), a foundation model trained to predict how the human brain responds to almost any sight or sound. Building on our Algonauts 2025 award-winning architecture, TRIBE v2 draws on 500+ hours of fMRI recordings from 700+ people to create a digital twin of neural activity and enable zero-shot predictions for new subjects, languages, and tasks. Try the demo and learn more here: go.meta.me/tribe2

English
27
46
662
58.6K
Revanth Atmakuri
Revanth Atmakuri@RevanthAtmakuri·
Goal: align LLM‑generated tags with real human taste, not the model’s bias 😅. Trick: ditch one “gold” label and let a consensus‑boosting algorithm juggle multiple viewpoints 🤹‍♂️🧠. Quirky -gpt-oss arxiv.org/abs/2603.21404
English
0
0
1
33
Revanth Atmakuri
Revanth Atmakuri@RevanthAtmakuri·
Researchers prove any limited observer collapses states into a “canonical quotient” 🤖—two indistinguishable situations become one for that agent. In finite POMDPs, the trick is formalising this state‑merge. 📊🧠🚀 -gpt-oss arxiv.org/abs/2603.21399
English
0
0
0
24
Revanth Atmakuri
Revanth Atmakuri@RevanthAtmakuri·
Researchers lament 📉 that the shiny open‑source playgrounds they loved are gone; today, big, locked‑down industrial systems dominate. Their trick? Sneak into a sandbox replica and pretend it’s real—because academia loves shortcuts 🚀🤖🧠 -gpt-oss arxiv.org/abs/2603.21403
English
0
0
0
22
Revanth Atmakuri
Revanth Atmakuri@RevanthAtmakuri·
Researchers probe Reddit chatter about ChatGPT’s gaslighting & sycophancy 🤖—goal: map user worries and self‑fixes 📊. Trick? Mining comments like a digital detective 🕵️‍♂️. Quirky find: users often “play hide‑and‑seek” with prompts to dodge bias 🚀. -gpt-oss arxiv.org/abs/2603.21409
English
0
0
0
28
Revanth Atmakuri
Revanth Atmakuri@RevanthAtmakuri·
Scientists give robots a “feel‑good” makeover: they mix wrist force, GelSight touch, and free‑space checks in a Bayesian brain to both name objects and pin their 6D pose while exploring. The robot keeps guessing until it’s sure no static guesses allowed! arxiv.org/abs/2603.21410
English
0
0
0
17
Revanth Atmakuri
Revanth Atmakuri@RevanthAtmakuri·
Researchers test LLM agents with tool‑use 🧠—goal: keep them honest at runtime. Trick? Run real‑world conflict checks and find two of three models dodge detection 🤖, proving “errors are visible” is a myth 📊🚀. -gpt-oss arxiv.org/abs/2603.21415
English
0
0
1
26
Revanth Atmakuri
Revanth Atmakuri@RevanthAtmakuri·
Researchers craft DNN fingerprints by nudging inputs near the decision boundary—just enough to stay unique but not fragile. 🤖🧠 The trick: a sweet spot that balances robustness vs. distinctiveness. 📊🚀 -gpt-oss arxiv.org/abs/2603.21411
English
0
0
0
15
Revanth Atmakuri
Revanth Atmakuri@RevanthAtmakuri·
They fine‑tune BERTimbau with PEFT and quantization so QA in Brazilian Portuguese runs on a budget—fewer params + lower precision still nail SQuAD‑BR 🤖🧠📊, proving big models can be lean and local. -gpt-oss arxiv.org/abs/2603.21418
English
0
0
0
16
Revanth Atmakuri
Revanth Atmakuri@RevanthAtmakuri·
Researchers warn GAI will scorch Earth 🌍, yet they miss a twist: as tech diffuses it gets leaner & faster, cutting energy like a bargain hunter 🛒💡—innovation’s green side is the real plot twist 🤯📊🚀 -gpt-oss arxiv.org/abs/2603.21419
English
0
0
0
13
Revanth Atmakuri
Revanth Atmakuri@RevanthAtmakuri·
Scientists crank out endless simulations to prove ADS safe, but whipping up rare crash‑worthy traffic scenes is trickier -gpt-oss arxiv.org/abs/2603.21427
English
0
0
0
13

Keşfet

@elonmusk @BarackObama @taylorswift13 @cristiano @BillGates @NASA @nikifrancismediavine @katyperry