Jonathan Balloch

I am a few days late on this, but I'm proud to say: I have passed my dissertation defense for my PhD! 🎉 I'm so thankful for my advisor @markriedl.bsky.social , my committee, my family, my lab, and all those who have supported me through this. Excited for this next chapter!

Le Chat underrated

well thats not great

I think a deeper difficulty in ML is the economy of attention. The hundreds of papers each day released on ArXiv in ML means that a reader needs to resort to heuristics to keep up. Stuff like trust a recommender system, or only read famous authors, or scan for buzzwords.

Sarah Paine is incredible

Given whats going on in the world, I think its time to reread Brave New World

Example, pre-train (reward free) to map temporal distances into distances in latent space, and then, finetune: map these through a dot product with a latent task description to a reward function. A couple of refs: openreview.net/forum?id=YGh... arxiv.org/abs/2110.02719 arxiv.org/abs/2110.15191

very exciting!

$14k open source humanoid robot upper torso. Writing with a pen on a notebook that you're holding is an impressively challenging task! Also comes with an open, modular, python software stack for robot control and planning. openpyro-a1.github.io

Hiring researchers and engineers for a stealth, applied research company with a focus on RL x foundation models. Folks on the team already are leading RL / learning researchers. If you think you'd be good at the research needed to get things working in practice, email me

Congrats Andrew and Rich, well deserved!! apnews.com/article/turi...

One reason to be intolerant of misleading hype in tech and science is that tolerating the small lies and deception is how you get tolerance of big lies

Trying to tell the story behind this explosion of research we are in. An unexpected RL Renaissance. New talk! Forecasting the Alpaca moment for reasoning models and why the new style of RL training is a far bigger deal than the emergence of RLHF. YouTube: https://buff.ly/41bVRPp

Easier installation, faster PPO script, new tutorials. The team has put in so much work and I'm excited for y'all to try it. github.com/Emerge-Lab/g...

Incredibly cool article. Why, in spite of all of the hype about the scale of learning, we shouldn't forget the second half of Sutton's Bitter Lesson: search scales too, and often better yellow-apartment-148.notion.site/AI-Search-Th... (h/t klowrey)

Awesome!!!

An appeaser is one who feeds a crocodile—hoping it will eat him last. —Winston Churchill

I try to avoid political topics but this is important. Trump NIH and FDA nominees debut new science journal "increasing transparency" of peer review, but more importantly giving a platform for science that can't pass traditional peer review www.foxnews.com/politics/tru... Stay skeptical my friends

incredible resource

This review paper by @guillaume-garrigos.com on SGD-related algorithms is a fantastic resource, offering elegant, self-contained, and concise proofs in a single, accessible reference. arxiv.org/pdf/2301.11235

As someone who has reported on AI for 7 years and covered China tech as well, I think the biggest lesson to be drawn from DeepSeek is the huge cracks it illustrates with the current dominant paradigm of AI development. A long thread. 1/

Its a bad time

@merve.bsky.social said it best👇

@markriedl.bsky.social 😶

#NeurIPS2024 wrapped up last week. I put together a curated reading list for #DeepRL and #reinforcementlearning work. (represents my interests). Talks and workshops: third-crowd-c77.notion.site/NeurIPS2024-... Curated reading list fracturedplane.notion.site/NeurIPS2024-... #Holidayreading

Since we are nearing the end of the year, I'll revisit some of our work I'm most excited about from the last year and maybe a sneak peek of what we are up to next. To start of, Robot Utility Models, which enables zero-shot deployment. In the video below, the robot hasnt seen these doors before.

If you’re in grad school, finding a therapist can be really helpful. The thing you’re doing is hard and it’s harder if you don’t have help managing imposter syndrome, stress, self esteem, and a whole bunch of other things.

This is why quantitative stock traders are always so stressed

Write your training loops with care, as if your ML model hated you and wanted you to fail.

My actual favorite niche starter pack here: go.bsky.app/2Gibu1a

If you want a version of DeepSeek-R1 or GPT-o1 that does a really good job on exactly one task, you can do that for ~$30 www.tomshardware.com/tech-industr... Which is to say that technical implementation of o1/o3/R1 style models are relatively easily achieved. Big tech AI companies have no moat.

The more you know

This seems to be maintained now: aideadlines.demelo.org

Challenge, starting now: no more undeciphered languages by 2030. This is an actually important thing to use #AI for. Please, my fellow AI and especially NLP people, if you agree, send this message to your network.

Hey @bsky.app can we get an option to filter feeds? as in by recency, engagement, etc?

Sometimes, if your imposter syndrome is in a place where it can be tested a little, its good just to head over to github and see what Phil Wang is working on github.com/lucidrains

Oh, what a great time to be entering the job market!!! 😅

this is the way

sneak peek at @ai2.bsky.social’s tulu 4