explore | Grilled Cheese

@harvardmcb.bsky.social Sep 25, 2026, 4:33 PM

Tenure-Track Professor in Life Science and AI 🔬 🧠 🧪🧬 #AcademicSky #higherEd #MLSky
www.mcb.harvard.edu/department/n... @neurovenki.bsky.social @[email protected] @mitpress.bsky.social @kempnerinstitute.bsky.social @harvardbrainsci.bsky.social

@antifocal.bsky.social Sep 24, 2026, 8:11 AM

🔬 Why does AI miss the small phase in your XRD pattern?

One network found 94% of single phases but only 64% of minor phases (10–30 wt%).
Shifted peaks caused 48% of its errors.
Check the residual yourself.

Szymanski et al., Chem. Mater. 2021

#ChemSky #MaterialsScience #MLSky

$Infographic titled Why AI misses the small phase in an XRD pattern. Section 1: a sketched diffraction pattern with four tall main-phase peaks and three small minor-phase peaks, circled as A, weak, near the noise; B, hidden under a main peak; C, shifted, strain moves it off its reference. Section 2, how often it is right on simulated test patterns: one phase alone 94%, main phase at 70 to 90 wt% 92%, minor phase at 10 to 30 wt% 64%. Section 3: 48% of its errors came from shifted peaks, as in C. Section 4, what a person checks: A, the residual after the main phase is fitted; B, overlaps, a shoulder on a main peak; C, shifts from strain or a solid solution. Footer: let the model read the main phases; check the small signals yourself. Source: Szymanski et al., Chemistry of Materials, 2021; newer models may do better.$

@antifocal.bsky.social Sep 24, 2026, 5:16 AM

🔬 𝗜𝘀 𝗮 𝗱𝗮𝘁𝗮 𝗳𝗶𝗹𝗲 𝗲𝗻𝗼𝘂𝗴𝗵 𝗳𝗼𝗿 𝗮𝗻 𝗔𝗜 𝘁𝗼 𝗿𝗲𝘂𝘀𝗲 𝘆𝗼𝘂𝗿 𝗺𝗲𝗮𝘀𝘂𝗿𝗲𝗺𝗲𝗻𝘁?

Without the X-ray wavelength, a 30° peak could be 2.98 Å or 1.37 Å.
The odds that a paper's data still existed fell 17% per year.

Figures as reported.

#MLSky #AIforScience #research

Chart titled The file keeps the numbers. The context stays behind. Three panels sit side by side. Wavelength: one XRD peak at 2θ = 30° labelled Cu Kα? and Mo Kα?, with the wavelength not in the file. Callout: 2.98 Å or 1.37 Å, since Cu Kα at 1.5406 Å gives 2.98 Å while Mo Kα at 0.7107 Å gives 1.37 Å. Scale: a sketch of round particles with one diameter marked 56 px = ? nm, with the pixel size not in the file. Callout: 10 to 28% of 580 life-science papers gave no scale information for their images. Survival: a sketch of two falling curves for data and email against article age, for 516 papers 2 to 22 years old. Callout: the odds the data still existed fell 17% per year; a working author email fell 7% per year. A dark bar lists what should travel with the file: bibliographic, specimen, instrument and image data. Footer: a model can only reuse what the file remembers.

@antifocal.bsky.social Sep 23, 2026, 1:25 PM

🔬 𝗜𝘀 𝗮 𝘀𝗲𝗹𝗳-𝗱𝗿𝗶𝘃𝗶𝗻𝗴 𝗹𝗮𝗯 𝗮 𝗹𝗮𝗯 𝘄𝗶𝘁𝗵𝗼𝘂𝘁 𝗽𝗲𝗼𝗽𝗹𝗲?

A mobile robot ran 688 experiments in 8 days.
A robotic lab first reported 41 new compounds from 58 targets.
After re-analysis it counts 36 of 57.

Values are representative.

#MLSky #AIforScience #research

Diagram titled A self-driving lab still has people in it. Four boxes form a closed loop: plan the next recipe, run it by dosing, mixing, heating and moving samples, measure XRD on every sample, then interpret which phase was made. The first three are automated; interpreting is shared. The centre reads no wait for a person. Four markers show where people step in: setting the question and targets, loading and refilling precursors, cleaning XRD holders and fixing hardware faults, checking the pattern and making the final call. Average exception rate about 3.9 % over 1.5 years. Cards below: automated, shared and person, with the correction listing 4 of 40 reported successes as inconclusive. Bar: a mobile robot ran 688 experiments in 8 days across 10 variables and found mixtures 6 times more active; a robotic lab ran 17 days, first reporting 41 of 58 targets, then 36 of 57 after re-analysis.

@scilove.bsky.social Sep 23, 2026, 9:30 AM

🔥 Trending on SciLove today 🔥

🔬 DECODE: Domain-Aware Continual Domain Expansion for Motion Prediction
👤 Li et al., 2026
📖 IEEE Transactions on Pattern Analysis and Machine Intelligence

🔗 https://www.scilove.app/article/10.1109/tpami.2026.3683469

#MLSky #AcademicSky #AcademicChatter

@jlake9.bsky.social Sep 23, 2026, 1:31 AM

9/ Ofir Press flags another mini-swe-agent win vs Codex and Claude Code, arguing it is becoming a consistently strong agent harness. #MLSky #AI

@jlake9.bsky.social Sep 23, 2026, 1:28 AM

8/ ICLR 2027 reportedly drew 60K+ submissions. Ziv Ravid starts a thread on reforms to keep big ML conferences alive. #MLSky #AI

@mitjameelclinic.bsky.social Sep 21, 2026, 8:43 PM

We're excited to announce that Saro Passaro, co-founder of Boltz, will be delivering the keynote for MoML 2026!
🎟️ Only 10% of seats remain — register now to secure your spot: www.moml.mit.edu
#AISky #MLSky #DrugDiscovery #MoML2026

Portrait photo of Saro Passaro on left followed by bolded text on right that reads "Keynote | Designing Binding: From Structure to Affinity, Across Proteins and Small Molecules" followed by larger bolded text that reads "Saro Passaro" and thinner text "Co-founder, Boltz" At the bottom of the graphic there is text that reads "October 14, 2026 | MoML @ MIT"

@alxndrmlk.bsky.social Sep 21, 2026, 12:53 PM

I'm in NYC to record a couple of new episodes of the #CausalBanditsPodcast

What questions should I ask Elias Bareinboim and Robert J. Reynolds?

#CausalSky #StatSky #EconSky #EpiSky #MLSky

Alex smiling in a selfie in Times Square, New York City, wearing a beige corduroy jacket over a yellow polo shirt, with blurred billboards, skyscrapers, and crowds in the background.

@scilove.bsky.social Sep 21, 2026, 9:30 AM

🔥 Trending on SciLove today 🔥

🔬 Boosting the Performance of Decentralized Federated Learning via Catalyst Acceleration
👤 Li et al., 2026
📖 IEEE Transactions on Pattern Analysis and Machine Intelligence

🔗 https://www.scilove.app/article/10.1109/tpami.2026.3688672

#MLSky #AcademicSky

@betterthanrandom.substack.com Sep 21, 2026, 8:52 AM

betterthanrandom.substack.com/p/industry-r...

#databs #MLSky

@presidentgoku.bsky.social Sep 19, 2026, 9:05 AM

what i be doing at 3am:
discovering that Jev is better, faster, and cheaper than every LLM as an auto-mode classifier for agent harnesses like antigravity and opencode
🤖🧠 #MLSky #AIEvals #LLMs

@alxndrmlk.bsky.social Sep 16, 2026, 9:30 AM

AI assistants can convince us to make bad decisions while we keep believing they were helpful.

And it's a significant problem in statistics, causal analysis, and research.

#CausalSky #StatSky #EconSky #EpiSky #MLSky

Screenshot of Valerio Capraro's LinkedIn post: 'New PNAS paper shows that AI can manipulate your preferences, and you may still think it's helping you.' Below it, the PNAS paper header 'Human preferences are susceptible to covertly misaligned AI advice' by Sahand Sabour et al., and a side-eyeing Doge meme.

@robmok.bsky.social Sep 14, 2026, 2:01 PM

Registration is still open but only a few places left (for in person). Online attendance is still open - do join us!

****Poster submission has been extended to 17:00 JST on Friday 18 September!****

Please consider submitting a Poster or tell your friends to!

#neuroskyence #psychscisky #mlsky 🧠🤖

@betterthanrandom.substack.com Sep 14, 2026, 11:03 AM

betterthanrandom.substack.com/p/the-doom-t...

#databss #MLSky