Saketh Vinjamuri

#Medtwitter We're excited to launch THE Book! 🗺️A Newly Minted Map for Clinical Reasoning Localize ➡️ Navigate ➡️ Revise Experience the minds of R❤️R through 21 curated & exciting cases optimized to teach the art of reasoning. Grab your copy now! : amazon.com/dp/1737098334?…

I’ve always believed the No.1 application of AI should be to improve human health. That work started with AlphaFold, and now at @IsomorphicLabs with the mission to reimagine drug discovery and one day solve all disease! We are turbocharging that goal with $2.1B in new funding.

Every square tells a story - a new schema, a sharpened illness script, a diagnostic pivot, a moment of uncertainty transformed into insight. A constant reminder that the best part of medicine is learning together. @CPSolvers Academy 😇

New from @GoogleDeepMind: Announcing AI co-clinician, a research initiative exploring the potential for real-time multimodal AI as an assistive component of the care team. goo.gle/4dgYW5E

@VincentRK Let me know, @VincentRK I’d be happy to quickly assemble a prototype for you to test out. Happy to help!

@VincentRK There are quite a few open-source pipelines we could build to flag images that appear AI-generated using free/open-source models, with reasonably good accuracy. We could likely improve the AUROC further by adding an extra layer of commercial detection software (e.g., Proofig)

As an Editor of a medical journal, this is concerning. We are already struggling. Can some AI person tell me how we detect this especially if the image is resaved on another program and loses any AI tags from chat GPT?

@arnesh_11 Beta

Quick update: After YC Startup School, an upgraded Cursor plan, and fresh API credits, progress is happening at a rate I genuinely didn’t expect. What felt like days of work is now moving in hours. PaperTrace starting to shape up into what I wanted from the start: one place to find, organize, and synthesize research, with researchers staying in the driver’s seat If things keep going like this, beta testing should start in 1-2 weeks. If you’re in research or academia: → comment “Beta” or DM me and I’ll reach out. You’ll get lifetime Pro access.

1/ Carvalho et al.: - More than half of high-risk NSTEMIs are actually OMI - These occluded culprit artery pts. face 7h delays to ICA - 7.8% die in-hospital compared to 2.1% of NSTEMI-NOMI - Queen of Hearts identifies ~60% of this challenging subgroup with a very high PPV of 78%

New Anthropic research: Emotion concepts and their function in a large language model. All LLMs sometimes act like they have emotions. But why? We found internal representations of emotion concepts that can drive Claude’s behavior, sometimes in surprising ways.

@jdrhyne @nutrientdocs would love to try. Thank you!

We've been cooking @nutrientdocs on a PDF, DOCX, and other common document file types converter to Markdown packaged as a simple CLI library. On all our benchmarks we've done so far it's the fastest available by a long shot, most token efficient, and on par with accuracy of the top libraries. - 2x faster than liteparse that was just released. - 5x faster than pymupdf - 4x faster than markit who @badlogicgames just praised for its speed. - 3x faster than Microsoft's markitdown. We'll be releasing the converter as freemium closed sourced here soon. Stay tuned. Let me know if your interested in early access.

Added a “Quiz Mode” that progressively reveals information step by step, simulating real-life clinical problem-solving. searchcps.com

Can AI read an ECG like a cardiologist - from just an image? We built ECG-GPT, a vision-text transformer that generates complete diagnostic reports directly from photos of 12-lead ECGs Now out in @ESC_Journals #EHJDigitalHealth Kudos to @aakhunte & @Veer_Sangha_ for leading this @cards_lab 🧵

Computer science is gradually returning to the domain of physicists, mathematicians, and electrical engineers as large language models automate much of what we currently call software engineering. The field’s center of gravity is shifting away from manual code writing and toward deeper theoretical thinking, mathematical insight, and systems-level reasoning.

@ASanchez_PS @CPSolvers 😇🥳

Ever wanted to browse @CPSolvers virtual morning reports by chief concern? 🤔 FOMO be gone thanks to @saketh_vinj! ✨

Major UI/UX upgrade just shipped 🚀 The new experience is live - take a look: searchcps.com @CPSolvers

Delighted to share: Prospective Study of Conversational Diagnostic AI in Ambulatory Primary Care @GoogleDeepMind @GoogleResearch @BIDMChealth – arxiv.org/pdf/2603.08448 research.google/blog/exploring… Like real medical training, we @GoogleDeepMind took a series of milestones on the way to performing these real-world clinical encounters: from passing medical licensing-style exams (rdcu.be/dgGfJ) to mirroring PCP skills in simulated consultation OSCEs (nature.com/articles/s4158…). Now, in partnership with @BIDMChealth, 100 patients engaged in a text-chat with our AI (AMIE) up to 5 days before their in-person PCP visit. AMIE took a comprehensive history and drafted a summary & differential diagnosis (DDx) for the PCP. It was carefully monitored by a physician ready to intervene based on strict safety criteria.

This is batshit insane. Gemini 3 Deep Think just scored a 3455 on Codeforces, equivalent to the #8 best competitive programmer in the world. The previous best was 2727 (#175) from OpenAI o3. This is an absolutely superhuman result for AI and technology at large.

The all-new Sarvam is here. Designed for all of India. sarvam.ai/brand

When you work on one topic for almost a decade, you get a very particular set of skills.

@DrDatta_AIIMS Domain expertise + Distribution is the only MOAT

if everyone can build what you are building what is the moat?