Caleb Darden

For those of you who, like me, refuse to pay for cable but want to watch March Madness... a new incognito window on Chrome (CMD+SHIFT+N) restarts the 1-hour preview

Squeak squadron boot camp

This has been a fun few weeks. Squeakview now includes: -multicamera (FLIR) for stereovision -Google drive backend via rclone because who likes pulling data -darker, sleeker, less clunky GUI -multiclass prediction -depth sensing via ZED* -stuff github.com/dlhagger/Squea…

@ylecun @grok who are the top investors in this?

Unveiling our new startup Advanced Machine Intelligence (AMI Labs). We just completed our seed round: $1.03B / 890M€, one the largest seeds ever, probably the largest for a European company. We're hiring! [the background image is the Veil Nebula - a picture I took from my backyard, most appropriate for an unveiling] More details here: techcrunch.com/2026/03/09/yan…

𝗪𝗵𝗮𝘁'𝘀 𝘁𝗵𝗲 𝗿𝗲𝗹𝗮𝘁𝗶𝗼𝗻𝘀𝗵𝗶𝗽 𝗯𝗲𝘁𝘄𝗲𝗲𝗻 𝗺𝗮𝗻𝗶𝗳𝗼𝗹𝗱𝘀 𝗮𝗻𝗱 𝗿𝗲𝗰𝘂𝗿𝗿𝗲𝗻𝘁 𝗻𝗲𝘁𝘄𝗼𝗿𝗸𝘀 𝗶𝗻 𝘁𝗵𝗲 𝗯𝗿𝗮𝗶𝗻? This looks like a must read (suppl material bursting with goodies). doi.org/10.1016/j.neur…

lol

This has gotten out of control

worth keeping this place alive givebutter.com/SaveRSC

@Laing_BrentonT Thank you!!! Feel free to message me or email David if you are interested in/have any questions about implementation!

@NeuroDarden This is really cool. Nice work

This is finally out. Cool stuff like model-assisted labeling, ultra (I mean ultra) fast training and inference. Models out of the box for MouseHouse. Real-time deployment on edge w/ NVIDIA Jetson. Oh, and 120 FPS real-time inference w/ quantized models! biorxiv.org/content/10.648…

@saen_dev PTQ via TensorRT. For FP16, its a direct post training weight cast. for int8 its PTQ again with a calibration set. Performance drop (MAP50-95/pixel error/etc) with FP16 is negligible. With INT8, you get some accuracy hits. We run 120 FPS on full 1440x1080 images with FP16 though!

@NeuroDarden 120 FPS real-time inference with quantized models on Jetson is insane. Model-assisted labeling saves so much time in the training loop. What's your quantization strategy - PTQ or QAT? And how much accuracy drop are you seeing with the quantized version?

This company is pushing every limit out there. What a time to be alive and getting ready for a career in science. techcrunch.com/2026/01/05/nvi…

"My jaw is on the floor" moments are rare. As a kid, they were on mountain tops. As an adult, they've come from science. In the last year, most have involved ML. This morning, it was watching a #yolo11 pose model crush it with some pretraining help from #DINOV3.

@zhuokaiz This is elite

Finally had time to prep intake materials for my eb-1a petition and realized I actually never had a proper academia-style CV 😅 Looked around, (surprisingly) didn't find many templates I liked, so I made one. Nothing too special, but I think it looks better, and has all the macros ready to use. Sharing it here in case it’s useful: github.com/zhuokaizhao/ac…

@techNmak LLM

These are literally the kind of LLM interview questions most candidates wish they had seen earlier. A curated list of LLM interview questions - shared by Hao Hoang Want this doc? Follow @techNmak and comment “LLM” - I’ll send it over.

I'm shocked by the power of desktop computing from NVIDIA. The GB10 trains pt models that previously took half a day on a 3060 in about a minute...

@NVIDIAGeForce GeForce Season

Only 200 Cyberpunk 2077 GPUs were ever created. Five years later, one returns. This original Cyberpunk 2077 GeForce RTX 2080 Ti now signed by Jensen Huang is back to celebrate five years in Night City🌆 Comment “GeForce Season” to enter.

Normalize naming your ipynb the question it’s supposed to answer.