彤彤

@fiddler_ai Good

Your teams are shipping AI faster than ever. Are they keeping up with observability, governance and trust? Fiddler is the Control Plane for enterprise AI. Giving teams standardized telemetry, continuous monitoring, and audit-ready governance across the agentic lifecycle.

@Eliana_work Wonderful woman!

@turingpi Build

🚨It’s Giveaway Time!🚨 Turing Pi 2.5 Cluster Board Giveaway Reply with "Build" to enter! Winner will be announced on January 27. Good luck!

@997unix @ollama @deepseek_ai Hello,how to set proxy on ollama?

@ollama @deepseek_ai literally: 'ollama run deepseek-r1:7b' on my M2/24GB Mac. It's fast and puts the "stream of reasoning" in the output. Very clean out-of-the-box, 10's of seconds of thinking time, then 80ish chars every 1-2 sec.

DeepSeek's first-generation reasoning models are achieving performance comparable to OpenAI's o1 across math, code, and reasoning tasks! Give it a try! 👇 7B distilled: ollama run deepseek-r1:7b More distilled sizes are available. 🧵

@kapehe_ok @vercel Beautiful

The day is finally here! Day 1 of new job! Head of Developer Community …. …. at @vercel 💃 I dreamt of triangles and v0 all last night! 🥹🥹🥹🥹🥹

@annadgoldie Thanks

这对我来说是个人事。我以前有机会在中国工作和教课，我永远忘不了有这么多人花时间与我交谈并帮助我学习语言。我非常感谢，也非常尊重中国人民。而且每次我提到我是麻省理工的，人们总是很友善和尊重。所以我必须说尊重和钦佩是相互的！无论国际或种族，我们都是人类兄弟姐妹，都值得被视为个体。

@HiFreyy 这个在国内可以激活吗

数字移民必备美国实体手机卡，ultramobile紫卡，每月3美元可以长期国内漫游，开wificalling收发短信免费，这款卡我已经稳定使用一年了，去年买的时候一张卡就要250块钱，刚才淘宝看了一下有一家店现在卖149，非常超值，强烈建议入手，真正的美国实体手机卡，非虚拟运营商，友情推荐，无任何利益相关，如果想选自己喜欢的号码，我明天教大家用5美元选自己喜欢的号码 traveldetail.fliggy.com/item.htm?&id=7…

@nutanlimaye Gpt will tell you

Can someone recommend a good textbook for teaching probability theory to data science students? It is for a course on discrete probability and statistics... For first year undergraduate data science students.

@AndrewLampinen @kareem_carr Good books

@kareem_carr Are you familiar with the PDP books (mitpress.mit.edu/9780262680530/…)? They focused on NNs as cognitive models more constrained by some aspects of neural processing, and also introduce concepts like representation learning by backpropagation which are rather fundamental to modern DL.

People in my mentions keep repeating that artificial neural networks were "inspired" by real neurons. I guess we are basing arguments on "inspiration" now. Are we in art class?

@LauraXiangMD Hello!

Hi #MedTwitter! Here to introduce myself with my first tweet. My name is Laura and I am a first-year dermatology resident at Cleveland Clinic. Looking forward to connecting with everyone! #DermTwitter

@JayathiMurthy @UCLA @UCLAengineering @OregonState Why leaving?

@kaur_simran25 If flatness does not, what matters?

Is flatness indicative of generalization? Not necessarily. Our experimental study calls the relationship between flatness (as measured by the max Hessian eigenvalue) and generalization into question. arxiv.org/abs/2206.10654

@sainingxie @AjdDavison Like the dark matter in your title!

@AjdDavison "Unsupervised learning is an ill-defined and misleading term that suggests that the learning uses no supervision at all. In fact, self-supervised learning is not unsupervised, as it uses far more feedback signals than standard supervised and RL methods do."ai.facebook.com/blog/self-supe…

Can anyone explain if there is a difference between unsupervised and self-supervised learning? To me they seem the same and I find myself using both terms interchangeably (I prefer unsupervised), but I feel like I'm confusing people who understand them to mean different things.

Become a @Discover Cardmember and you'll get a $100 Statement Credit with your 1st purchase within 3 months. refer.discover.com/s/wangtnetlab

@cvxpy_team @MaxSchaller8611 In cvxpy, I have an infeasible starting point and the program can still get a result. Is it reasonable? Can anyone tell me about this? Thanks.

CVXPY now supports code generation, via the CVXPYgen package by @maxschaller8611: code: github.com/cvxgrp/cvxpygen paper: web.stanford.edu/~boyd/papers/c… CVXPYgen lets you solve parametrized CVXPY problems anywhere you can run C programs --- like on a drone.

@usuallyuseless @cgarciae88 BTW, would you tell me why should I choose jax instead of tensorflow? Thanks.

@cgarciae88 @wangtnetlab Also: automatic differentiation.

JAX code I ❤️ #2 In the old days you could code a nice pairwise formula but vectorizing added a lot of unpleasant artifacts 😕 (tiling, broadcasting). Using a double vmap however, you can teach your beautiful function to operate over sets without changing a single line 🔥

@usuallyuseless @cgarciae88 Thank you. I think this is the reason I should give jax a try instead of numpy.

@cgarciae88 Thanks for the answer!

@wangtnetlab You can use jax.jit to optimize your CPU code, as well as vmap, scan, and other utils to make your code cleaner /more efficient.

@cgarciae88 Thanks, so why we choose jax instead of numpy if we only run on cpu?

@wangtnetlab JAX runs on CPU, GPU, and TPU :)