Pierre-Antoine Bannier (@el_pa_b) 's Twitter Profile
Pierre-Antoine Bannier

@el_pa_b

Deep Learning & Bio @OwkinScience . Computational Pathology | Scientific Computing | Open Source. Graduate of @Polytechnique and @HECParis.

ID: 556281450

linkhttps://github.com/pabannier calendar_today17-04-2012 18:57:33

459 Tweet

755 Takipçi

293 Takip Edilen

Rafa Nadal (@rafaelnadal) 's Twitter Profile Photo

Mil gracias a todos Many thanks to all Merci beaucoup à tous Grazie mille à tutti 谢谢大家 شكرا لكم جميعا תודה לכולכם Obrigado a todos Vielen Dank euch allen Tack alla Хвала свима Gràcies a tots

Markus Eckstein (@markuseckstein3) 's Twitter Profile Photo

Erdafitinib is just around the corner for #metastatic #urothelialcancer! However, FGFR3 alterations are rare (10-15%) in MUC! In our recently published study in Nature Communications we thus asked whether we could use #AI to prescreen histopathology slides for presence of FGFR3

Erdafitinib is just around the corner for #metastatic #urothelialcancer! 

However, FGFR3 alterations are rare (10-15%) in MUC!

In our recently published study in <a href="/NatureComms/">Nature Communications</a> we thus asked whether we could use #AI to prescreen histopathology slides for presence of FGFR3
Markus Eckstein (@markuseckstein3) 's Twitter Profile Photo

1/10 The presented deep-learning model can identify FGFR3 mutations (important for targeted therapy) from routine H&E slides of muscle-invasive bladder cancer (MIBC). Here are the key points—and some caveats. 2/10 Study Design •Trained on 1222 cases from multiple cohorts

Pierre-Antoine Bannier (@el_pa_b) 's Twitter Profile Photo

NVIDIA integrated DeepSeek R1 to find new optimized CUDA kernels for different flavours of attention! developer.nvidia.com/blog/automatin…

Pierre-Antoine Bannier (@el_pa_b) 's Twitter Profile Photo

Researchers at Google have developed an AI co-scientist system built on Gemini 2.0 that uses multiple AI agents to generate, debate, and evolve scientific hypotheses. The system has shown promising results in biomedical applications, successfully identifying new drug candidates

Hamel Husain (@hamelhusain) 's Twitter Profile Photo

My thesis is LLMs significantly boost your motivation to work on something. Getting a shitty prototype fast is immensely motivating Code and time are often not the bottleneck, motivation is the bottleneck.

Pierre-Antoine Bannier (@el_pa_b) 's Twitter Profile Photo

Just tried (cc Henrik Anton Osika – eu/acc ) for the first time, and I'm in awe 😮 Two prompts and it created a neat 3d vector field visualizer with three.js. Crazy how smooth the experience is! You are now a few prompts away from building very cool stuff. Wild...

PJ Ace (@pjaccetturo) 's Twitter Profile Photo

What if Studio Ghibli directed Lord of the Rings? I spent $250 in Kling credits and 9 hours re-editing the Fellowship trailer to bring that vision to life—and I’ll show you exactly how I did it 👇🏼

Flo Crivello (@altimor) 's Twitter Profile Photo

People are often shocked to learn that *on average,* someone's net salary in France is 40% of what was paid by the employer (and this is not counting the 20% sales tax slapped on everything). This does add up: France's public spending hit 62% of GDP in 2020: the public sector is

JFPuget 🇺🇦🇨🇦🇬🇱 (@jfpuget) 's Twitter Profile Photo

My golden role when using python. Use python bindings to run code written in efficient languages like C, C++, CUDA. If you really have to write algorithmic code in python, then use relevant arrays (e.g. numpy, cupy, etc) and compile it with numba.

Pall Melsted (@pmelsted) 's Twitter Profile Photo

Excited to share this preprint that describes my latest work on using GPUs to accelerate processing of RNA-seq data. The title says it all: "RNA-seq analysis in seconds using GPUs" now on biorxiv biorxiv.org/content/10.648… Figure 1 shows they key result

Excited to share this preprint that describes my latest work on using GPUs to accelerate processing of RNA-seq data. 

The title says it all: "RNA-seq analysis in seconds using GPUs" now on biorxiv biorxiv.org/content/10.648… 

Figure 1 shows they key result