@spatial @spatialneuron

gradient learner github.com/dmbernaal sf/nyc Joined May 2019

Tweets

2K
Followers

243
Following

317
Likes

4K

Kendra LA Oct 14th-19th @StretchGoat_

2 days ago

1 1 16 561 0

Download Image

Deedy @deedydas

a week ago

This new DeepMind research shows just how broken vector search is. Turns out some docs in your index are theoretically incapable of being retrieved by vector search, given a certain dimension count of the embedding. Plain old BM25 from 1994 outperforms it on recall. 1/4

93 404 4K 465K 5K

Download Image

Binyuan Hui @huybery

2 weeks ago

I believe LLMs will inevitably surpass humans in coding. Let us think about how humans actually learn to code. Human learning of coding has two stages. First comes memorization and imitation: learning syntax and copying good projects. Then comes trial and error: writing code,…

125 115 1K 120K 324

Mr 9to5 @9to5Balance

2 weeks ago

Researchers, don’t miss this: ‘The Big LLM Architecture Comparison’ by @rasbt lays out how modern models like DeepSeek-V3 and Kimi K2 differ in structure, efficiency, and capabilities. Great for model design inspiration! Link in comments.

2 49 501 29K 493

Download Image

jack @jack

2 weeks ago

i love deleting code

958 764 11K 1.2M 408

DataVoid @DataPlusEngine

3 weeks ago

Just came up with Multi-Scale Control for Stable Diffusion and I'm losing my mind! Instead of your prompt flowing through ALL upsample/downsample blocks like normal, you can now inject DIFFERENT prompts at different resolution stages of the UNet. Discovered something wild:…

7 4 30 1K 15

Download Image

mac @connormclarenn

3 weeks ago

pressure is a crazy feeling it's an energy that will turn you into a pussy or a killer you either run through the fucking wall and build confidence OR shut down and feel bad for yourself BUT you can always get back up and run through the fucking wall SO the real question…

5 1 58 3K 13

Beff – e/acc @BasedBeffJezos

3 weeks ago

If you don't sacrifice literally everything in your life to the altar of the mission, the universe will punish you

50 32 402 20K 61

TotientQuotient @t0tientqu0tient

3 weeks ago

"math isn't a hard subject" math:

laine @osakasataandagi

3 weeks ago

"math isn't a hard subject" math: https://t.co/GSZ84Q9UcN

246 441 4K 490K 247

62 209 3K 139K 638

Download Image

Steven Schwartz @cultured

3 weeks ago

Whop is now officially independent from Stripe. What this means for anybody running a business on Whop: > Orchestration through multi-PSPs is now available (lifts your revenue by 5 - 10%) > Sellers can get paid out in nearly every country via local banks, BTC, or stablecoins…

371 202 2K 869K 792

Download Video

Tanishq Mathew Abraham, Ph.D. @iScienceLuvr

3 weeks ago

Time Is a Feature: Exploiting Temporal Dynamics in Diffusion Language Models "Our work here reveals a critical phenomenon, temporal oscillation, where correct answers often emerge in the middle process, but are overwritten in later denoising steps. To address this issue, we…

6 45 223 16K 137

Download Image

Alessandro Strumia @AlessandroStru4

4 weeks ago

The best logical explanation of entropy. (It only took me 11 months to find an hour to read it). arxiv.org/abs/2409.09232

26 318 3K 173K 4K

Tanishq Mathew Abraham, Ph.D. @iScienceLuvr

4 weeks ago

MolmoAct: Action Reasoning Models that can Reason in Space "Reasoning is central to purposeful action, yet most robotic foundation models map perception and instructions directly to control, which limits adaptability, generalization, and semantic grounding. We introduce…

3 19 164 13K 104

Download Image

Frank Nielsen @FrnkNlsn

4 weeks ago

Information geometry for deep learning: neuromanifold of stochastic NNs, Fisher-Rao norm, natural gradient, singularities. Relative Fisher information (RFIM) and relative natural gradient for tackling large NNs (ICML): proceedings.mlr.press/v70/sun17b.html

3 142 970 62K 788

Download Image

mitsuri @0xmitsurii

a month ago

"Study hard what interests you the most in the most undisciplined, irreverent and original manner possible." - Richard Feynman

3 19 163 5K 43

Download Image

mitsuri @0xmitsurii

a month ago

Larry Ellison: Being different is the only way to get ahead.

51 817 8K 130K 4K

Download Video

steve hsu @hsu_steve

a month ago

Is Chain-of-Thought Reasoning of LLMs a Mirage? ... Our results reveal that CoT reasoning is a brittle mirage that vanishes when it is pushed beyond training distributions. This work offers a deeper understanding of why and when CoT reasoning fails, emphasizing the ongoing…

200 970 6K 782K 5K

Download Image

Haider. @slow_developer

a month ago

Yann LeCun was right about LLMs. "we will never achieve AGI by simply scaling them up" LLMs will be a key part of achieving it -- but calling them alone AGI is not enough we need at least one major architectural breakthrough, if not many so it's up to the labs now how quickly…

196 141 2K 183K 370

Jia-Bin Huang @jbhuang0604

a month ago

Diffusion LLMs are promising ways to overcome the limitations of autoregressive LLMs. Less error propagation, easier to control, and faster to sample! But how do Diffusion LLMs actually work? 🤔 Let's explore some ideas on this fascinating topic! youtu.be/8BTOoc0yDVA

18 109 769 62K 594

Download Video

François Fleuret @francoisfleuret

a month ago

Guys, I understand you like drama, but this is a remark about the AI development at large. We are seeing the plateau: just scaling up is coming to an end. For EVERYONE, not one company in particular.

François Fleuret @francoisfleuret

a month ago

Guys, I understand you like drama, but this is a remark about the AI development at large. We are seeing the plateau: just scaling up is coming to an end. For EVERYONE, not one company in particular.