Alessandro Crimi ๐Ÿง ๐Ÿงฌ๐Ÿ”ฌ๐Ÿฉบ( @alecrimi.bsky.social ) (@dr_alex_crimi) 's Twitter Profile
Alessandro Crimi ๐Ÿง ๐Ÿงฌ๐Ÿ”ฌ๐Ÿฉบ( @alecrimi.bsky.social )

@dr_alex_crimi

Prof.@AGH_Krakow. Intersecting #neuroscience/#AI and #impact #social development.
#innovation #globalhealth All links: linktr.ee/alecrimi

ID: 411753065

linkhttps://www.alessandrocrimi.com calendar_today13-11-2011 20:47:43

5,5K Tweet

7,7K Followers

3,3K Following

Brain&More lab @AGH_Krakow (@brain_more) 's Twitter Profile Photo

"From knowledge generation to knowledge verification: examining the hashtag#biomedical generative capabilities of #ChatGPT" sciencedirect.com/science/articlโ€ฆ #ai #llm

"From knowledge generation to knowledge verification: examining the hashtag#biomedical generative capabilities of #ChatGPT"  
sciencedirect.com/science/articlโ€ฆ 
#ai #llm
Alessandro Crimi ๐Ÿง ๐Ÿงฌ๐Ÿ”ฌ๐Ÿฉบ( @alecrimi.bsky.social ) (@dr_alex_crimi) 's Twitter Profile Photo

Harvard & Roche just dropped COMPASS, interpretable foundation #AI model to predict #immunotherapy response across 33 #cancers. It beats 22 models predicting survival outcomes more accurately than PD-L1 expression, with a hazard ratio of 4.7 (p < 0.0001) medrxiv.org/content/10.110โ€ฆ

Harvard &amp; Roche just dropped COMPASS, interpretable foundation #AI model to predict #immunotherapy response across 33 #cancers.
It beats 22 models predicting survival outcomes more accurately than PD-L1 expression, with a hazard ratio of 4.7 (p &lt; 0.0001) medrxiv.org/content/10.110โ€ฆ
Alessandro Crimi ๐Ÿง ๐Ÿงฌ๐Ÿ”ฌ๐Ÿฉบ( @alecrimi.bsky.social ) (@dr_alex_crimi) 's Twitter Profile Photo

The largest-ever study of how radiologists collaborate with #AI just arrived, image and #llm ๐—ง๐—ต๐—ฒ ๐—–๐—ผ๐—น๐—น๐—ฎ๐—ฏ-๐—–๐—ซ๐—ฅ ๐—ฑ๐—ฎ๐˜๐—ฎ๐˜€๐—ฒ๐˜ ๐—ฐ๐—ฎ๐—ฝ๐˜๐˜‚๐—ฟ๐—ฒ๐˜€ over 73,000 radiologist-#pathology predictions from 227 radiologists and 104 thoracic pathologies. nature.com/articles/s4159โ€ฆ

The largest-ever study of how radiologists collaborate with #AI just arrived, image and #llm
๐—ง๐—ต๐—ฒ ๐—–๐—ผ๐—น๐—น๐—ฎ๐—ฏ-๐—–๐—ซ๐—ฅ ๐—ฑ๐—ฎ๐˜๐—ฎ๐˜€๐—ฒ๐˜ ๐—ฐ๐—ฎ๐—ฝ๐˜๐˜‚๐—ฟ๐—ฒ๐˜€ over 73,000 radiologist-#pathology predictions from 227 radiologists  and 104 thoracic pathologies.
nature.com/articles/s4159โ€ฆ
Alessandro Crimi ๐Ÿง ๐Ÿงฌ๐Ÿ”ฌ๐Ÿฉบ( @alecrimi.bsky.social ) (@dr_alex_crimi) 's Twitter Profile Photo

Hoping to help from #scizophrenia to #stroke , here is our "Investigating the interaction between #EEG and #fNIRS: A multimodal network analysis of brain connectivity" sciencedirect.com/science/articlโ€ฆ

Alessandro Crimi ๐Ÿง ๐Ÿงฌ๐Ÿ”ฌ๐Ÿฉบ( @alecrimi.bsky.social ) (@dr_alex_crimi) 's Twitter Profile Photo

BioReason: Incentivizing Multimodal Biological Reasoning within a #DNA-#LLM Model arxiv.org/abs/2505.23579 ๐Ÿงฌ it integrates DNA foundation models (Evo2) with LLMs connecting different biological knowledge like a scientist #AI #MachineLearning #genomics #biology #XAI #genomics

BioReason: Incentivizing Multimodal Biological Reasoning within a #DNA-#LLM Model 
arxiv.org/abs/2505.23579 
๐Ÿงฌ it integrates DNA foundation models (Evo2) with LLMs connecting different biological knowledge like a scientist

#AI #MachineLearning #genomics #biology #XAI
#genomics
Alessandro Crimi ๐Ÿง ๐Ÿงฌ๐Ÿ”ฌ๐Ÿฉบ( @alecrimi.bsky.social ) (@dr_alex_crimi) 's Twitter Profile Photo

watching spacetime curve in real time. Itโ€™s not just code. Itโ€™s a window into #dynamicsystems, entropy, and the fragile balance between order and collapse. Are we debugging the cosmosโ€ฆ or ourselves? #Cplusplus #GPU #BlackHole #ChaosTheory

Alessandro Crimi ๐Ÿง ๐Ÿงฌ๐Ÿ”ฌ๐Ÿฉบ( @alecrimi.bsky.social ) (@dr_alex_crimi) 's Twitter Profile Photo

'Inside you are many wolves: Using cognitive models to interpret value trade-offs in LLMs' Cool idea about #empathy, with cringe title :-) arxiv.org/abs/2506.20666 #AI #computation #language #ReinforcementLearning #EmergentBehavior #LLM #Socialreasoning #alignment

'Inside you are many wolves: Using cognitive models to interpret value trade-offs in LLMs'
Cool idea about #empathy, with cringe title :-)

arxiv.org/abs/2506.20666 

#AI #computation #language #ReinforcementLearning
#EmergentBehavior #LLM #Socialreasoning #alignment