🔥 Trending on SciLove today 🔥

🔬 Transforming Science with Large Language Models: A Survey on AI-assisted Scientific Discovery, Experimentation, Content Generation, and Evaluation
👤 Eger et al., 2026
📖 ACM Computing Surveys

🔗 https://www.scilove.app/article/10.1145/3845596

#MLSky #AcademicSky

Bubbly hyperbole is a lot more fun the dismal realities of revenue, EBITDA, and returning value to shareholders – but we ought not forget the bill will come due, and that might affect the tools in development and freely/easily accessible at present.
#medsky #MLsky

jamanetwork.com/journals/jam...

Networks trained on simulated images needed retraining on labelled real ones, about 30 min on one GPU. People set the drift limit, steered the models and ended each run.

Roccapriore et al., Science Advances 10, eadn5899 (2024)

Values are representative.

#MLSky #AIforScience #research

Errors from X and Y coils, never varied, were corrected least well. Very large distortions stay partly uncorrected. People ran the calibration and remove big solvent peaks first.

Jopa et al., J. Phys. Chem. B 129, 6078 (2025)

Values are representative.

#MLSky #AIforScience #research

betterthanrandom.substack.com/p/charitymax...

#databs #MLSky

Is causality enough to take AI to the next level?

And why do today's models fall so easily into a correlational black hole?

1/

#CausalSky #MLSky #StatSky #EpiSky #EconSky

We just sent today's issue of Causal Python Bi-Weekly.

Here's what's inside:

1/

#CausalSky #StatSky #EconSky #EpiSky #MLSky

Pulling back the veil on Epic again – imagine needing to activate their sepsis alert tools to get Gold Stars or join the Honor Roll. 🤮
#medsky #MLsky
www.evidencetriage.com/p/leave-the-...

Errors grew where training materials were sparse, worst near a refractive index of 1.58. Noise and real measurement variation are not yet built in. People set the parameter ranges.

Jung et al., Nanophotonics 14, 471 (2025)

Values are representative.

#MLSky #AIforScience #research

You can see me here citing Sam Altman's 2024 prediction about 1-person $1 billion companies created using AI.

1/

#CausalSky #StatSky #MLSky #AISky

📻 Available for Broadcast on PRX
exchange.prx.org/p/637604

#AI #AIAgents #MLSky #SciencePodcast #Heliox

When neighbouring elements scatter X-rays alike, ordered and disordered phases give near-identical powder patterns. People chose the targets and in the end re-analysed the patterns by hand.

Szymanski et al., Nature 624, 86 (2023)

Values are representative.

#MLSky #AIforScience #research

Patchy films gave noisy readings, so each sample ran two to four times. People set the ranges, picked the scale-up recipe and saw a DMSO test film form defects at 100 cm².

Wang et al., Nature Communications 16, 1498 (2025)

Values are representative.

#MLSky #AIforScience #research

400+ nodes and 1500+ edges

And 11 different ways people at NASA referred to changes in gravitational force

1/

#CausalSky #StatSky #EpiSky #MLSky

Two Bridge2AI Investigators receive the IEEE Innovation in Societal Infrastructure Award 🏆 Read more about our Investigators awarded for their leadership in privacy-preserving clinical research: bit.ly/4doccGq

#ArtificialIntelligence #BiomedicalInformatics #MedSky #MedAI #AISky #MLSky

When neighbouring elements scatter X-rays alike, ordered and disordered phases give near-identical powder patterns. People chose the targets and in the end re-analysed the patterns by hand.

Szymanski et al., Nature 624, 86 (2023)

Values are representative.

#MLSky #AIforScience #research

betterthanrandom.substack.com/p/do-ai-assi...

#databs #MLSky

People set the safe voltage range, chose the two plunger gates, designed the score and labelled every map. This was coarse tuning, not a qubit, with five runs per device.

Moon et al., Nature Communications 11, 4161 (2020)

Values are representative.

#MLSky #AIforScience #research

People still polish the surface, set voltage, tilt and detector distance, then check any phase the model has not seen.

Kaufmann et al., Science 367, 564 (2020)

Values are representative.

#MLSky #AIforScience #research

The detector loses the phase, so different profiles can fit one curve. The operator set bounds for every parameter. In a PTCDI-C₈ test, results below about 250 Å scattered more.

Pithan et al., J. Synchrotron Rad. 30, 1064 (2023)

Values are representative.

#MLSky #AIforScience #research