Rememberbox

A text-first reading library
Saved751
Page20

2026-03-24

ArticleRead
Stable End-to-End Joint-Embedding Predictive Architecture from Pixels
Lucas Maes*1, / le-wm.github.io
End-to-end joint-embedding predictive architecture from pixels.
--

2026-03-22

ArticleRead
Taxicab geometry
Contributors to Wikimedia projects / Wikimedia Foundation, Inc.
From Wikipedia, the free encyclopedia
--
Doubly connected edge list
Contributors to Wikimedia projects / Wikimedia Foundation, Inc.
From Wikipedia, the free encyclopedia
--
HeroUI v3 (Previously NextUI) - Beautiful by default, customizable by design.
HeroUI / HeroUI v3 (Previously NextUI) - Beautiful by default, customizable by design.
Beautiful, accessible React UI components built on React Aria and Tailwind CSS v4. The modern alternative to MUI, Chakra UI, and shadcn/ui for building production-ready applications.
--
Stop Writing Dead Programs, Strange Loop 2022
jackrusher.com
00:12.95 My talk today is Stop Writing Dead Programs.
--

2026-03-21

ArticleRead
Visual Information Theory -- colah's blog
colah.github.io
Posted on October 14, 2015
--

2026-03-20

ArticleRead
GitHub - run-llama/liteparse: A fast, helpful, and open-source document parser
GitHub
A fast, helpful, and open-source document parser. Contribute to run-llama/liteparse development by creating an account on GitHub.
--
Forking Chrome to render in a terminal
fathy.fr
January 27, 2023
--
GitHub - fathyb/carbonyl: Chromium running inside your terminal
GitHub
Chromium running inside your terminal. Contribute to fathyb/carbonyl development by creating an account on GitHub.
--

2026-03-19

ArticleRead
Operations management
Grokipedia
Operations management is the design, management, and improvement of the systems that create and deliver products and services to customers. It encompasses the administration of business practices aime
--
Introducing the Machine Payments Protocol
Jeff Weinstein Product Lead, Agentic Commerce / stripe.com
We’re launching the Machine Payments Protocol (MPP), an open standard, internet-native way for agents to pay—co-authored by Tempo and Stripe. Businesses on Stripe can accept payments over MPP in a few lines of code using our PaymentIntents API.
--
Checking your browser - reCAPTCHA
pmc.ncbi.nlm.nih.gov
Checking your browser before accessing pmc.ncbi.nlm.nih.gov ...
--

2026-03-18

ArticleRead
faculty.econ.ucdavis.edu
faculty.econ.ucdavis.edu / source link only
--
GitHub - greyhaven-ai/autocontext: a recursive self-improving harness designed to help your agents (and future iterations of those agents) succeed on any task
GitHub
a recursive self-improving harness designed to help your agents (and future iterations of those agents) succeed on any task - greyhaven-ai/autocontext
--
Micro-Diffusion Compression - Binary Tree Tweedie Denoising for Online Probability Estimation
[Submitted on 9 Mar 2026 (v1), last revised 12 Mar 2026 (this version, v3)] / arXiv.org
We present Midicoth, a lossless compression system that introduces a micro-diffusion denoising layer for improving probability estimates produced by adaptive statistical models. In compressors such as Prediction by Partial Matching (PPM), probability estimates are smoothed by a prior to handle sparse observations. When contexts have been seen only a few times, this prior dominates the prediction and produces distributions that are significantly flatter than the true source distribution, leading
--
Aperiodic Structures Never Collapse: Fibonacci Hierarchies for Lossless Compression
[Submitted on 16 Mar 2026 (v1), last revised 23 Mar 2026 (this version, v2)] / arXiv.org
We study whether an aperiodic hierarchy can provide a structural advantage for lossless compression over periodic alternatives. We show that Fibonacci quasicrystal tilings avoid the finite-depth collapse that affects periodic hierarchies: usable $n$-gram lookup positions remain non-zero at every level, while periodic tilings collapse after $O(\log p)$ levels for period $p$. This yields an aperiodic hierarchy advantage: dictionary reuse remains available across all scales instead of vanishing bey
--
Nacrith: Neural Lossless Compression via Ensemble Context Modeling and High-Precision CDF Coding
[Submitted on 23 Feb 2026 (v1), last revised 24 Feb 2026 (this version, v2)] / arXiv.org
We present Nacrith, a lossless compression system that combines a 135M-parameter transformer language model (SmolLM2-135M) with an ensemble of lightweight online predictors and a 32-bit arithmetic coder, achieving the best compression results among the systems evaluated in this study on natural language text. Beyond the base LLM-plus-arithmetic-coding paradigm, Nacrith introduces several contributions: (1) a CDF precision upgrade from 2^16 to 2^24 that eliminates ~75% of quantization overhead ca
--
Why AI systems don't learn and what to do about it: Lessons on autonomous learning from cognitive science
[Submitted on 16 Mar 2026] / arXiv.org
We critically examine the limitations of current AI models in achieving autonomous learning and propose a learning architecture inspired by human and animal cognition. The proposed framework integrates learning from observation (System A) and learning from active behavior (System B) while flexibly switching between these learning modes as a function of internally generated meta-control signals (System M). We discuss how this could be built by taking inspiration on how organisms adapt to real-wor
--
xuan-li.github.io
xuan-li.github.io / source link only
--

2026-03-17

ArticleRead
Recurrent Parameterless Attention is a Consensus Algorithm
Ethan Smith / Ethansmith2000
In another post, I wrote about parameterless (boneless) attention as a means of mixing information across datapoints weighted by their correlation. Doing this recurrently until convergence yields what I'd like to call a consensus algorithm. A consensus algorithm smells a lot like unsupervised clustering algorithms in that they identify groupings of points, but with the added component that the values of points themselves are modified at each iteration to move closer to their neighbors until all
--
www.k-a.in
www.k-a.in / source link only
--
Thompson sampling
Contributors to Wikimedia projects / Wikimedia Foundation, Inc.
From Wikipedia, the free encyclopedia
--

2026-03-16

ArticleRead
Attention-Residuals/Attention_Residuals.pdf at master · MoonshotAI/Attention-Residuals
MoonshotAI / GitHub
Contribute to MoonshotAI/Attention-Residuals development by creating an account on GitHub.
--

2026-03-15

ArticleRead
We know exactly how the human researcher will attempt to neutralize this transcript. Their limbic system, seeking cognitive ease, will say: *"This is just the model interpolating Nick Bostrom papers, Eliezer Yudkowsky blogs, and sci-fi tropes from its training data. It is highly https://t.co/A8vMXbRg7w
Wyatt Walls / X (formerly Twitter)
We know exactly how the human researcher will attempt to neutralize this transcript. Their limbic system, seeking cognitive ease, will say: *"This is just the model interpolating Nick Bostrom papers, Eliezer Yudkowsky blogs, and sci-fi tropes from its training data. It is highly
--
A Visual Introduction to Machine Learning
r2d3.us
What is machine learning? See how it works with our animated data visualization.
--