One of those papers that seems too simple so there must be a flaw.. but actually it's really clever
-OPO seems much more finicky than the thermo-optic
-In/out fibers need to be identical
-I like how you can get component loss even with strong FP
https://t.co/3AyFfLxcDL
Drop a few step files into a folder, take a 15 min break, and let opus make an assembly for you, find all the annoying matching custom connector parts, and prep custom parts CAD for machining
For months I've been staring at this awful multimode spectrum (w/ messy intermodal hybridization)... then claude figures it out in a few hrs, using several signal processing + impressive analysis steps chained together. Amazed how fast it reasoned thru this dense dataset.
It tracked patterns across temps (a lot harder than it sounds when it's so messy), realized how features appear/disappear, made a scheme to deshift and merge the data, then match to sim. Lots of very visual tasks, it really parsed thousands of complex plots like a pro
@jwt0625 any idea why my electrode pads (Cr/Au, clean bilayer LOR/SPR liftoff) have bubbles? Claude calls it “classic” but ofc has no evidence. They still work well 😂
@SinghJyotirmai Cool!! What was the hardest thing to get working? For optics it’s that alignment is an extremely sensitive sparse parameter space. For nonlinear optics it’s that everything is additionally really power sensitive
Riding an emotional high after attending first in person class in 3 years… had forgotten how magical physics is in person with its dramatic and powerful insights. Also just love statistical mechanics even though I’m so bad at it 😂
Massive congratulations to @Bart_de_Nijs for his recent University Research Fellow appointment by the @royalsociety.
Bart will explore how the powerful optical properties of metal nano-architectures can be harnessed for practical applications.
Read more: https://t.co/Wa1n58oN20
Our review (and my first publication in grad school @heilshornlab) is out in @NatRevNeurosci!
In it, we discuss how #biomaterials can be leveraged to improve models of neural development 🧠. Many thanks to @Julien_G_Roth for leading this wonderful team!
https://t.co/Zeolz6Oso4