che_shr_cat

che_shr_cat on Hacker News

Universal Transformers Need Memory: Depth-State Trade-Offs in Adaptive Recursive

Posted by che_shr_cat on Tuesday April 28, 2026 @01:21PM
from the arxiv.org dept.

https://arxiv.org/abs/2604.21999

( Read More... | no comments | on HN )

Game: Storm Cloud Simulator / Pure HTML5

Posted by che_shr_cat on Monday March 23, 2026 @09:59AM
from the cloud-heavy-industries.com dept.

https://cloud-heavy-industries.com/grumbulus/

( Read More... | no comments | on HN )

The Selfish Ribosome

Posted by che_shr_cat on Sunday March 08, 2026 @01:29PM
from the biorxiviq.substack.com dept.

https://biorxiviq.substack.com/p/the-selfish-ribosome

( Read More... | no comments | on HN )

Embedded Universal Predictive Intelligence: a coherent framework for multi-agent

Posted by che_shr_cat on Sunday December 07, 2025 @03:46PM
from the arxiviq.substack.com dept.

https://arxiviq.substack.com/p/embedded-universal-predictive-intelligence

( Read More... | no comments | on HN )

NeurIPS 2025 Best Papers in Comics: From Artificial Hivemind to 1000-Layer RL

Posted by che_shr_cat on Monday December 01, 2025 @11:11AM
from the gonzoml.substack.com dept.

https://gonzoml.substack.com/p/neurips-2025-best-papers-in-comics

( Read More... | no comments | 3 points | on HN )

Visualizing Research: How I Use Gemini 3.0 to Turn Papers into Comics

Posted by che_shr_cat on Monday November 24, 2025 @05:08PM
from the gonzoml.substack.com dept.

https://gonzoml.substack.com/p/visualizing-research-how-i-use-gemini

( Read More... | no comments | on HN )

Arc Is a Vision Problem

Posted by che_shr_cat on Saturday November 22, 2025 @08:37AM
from the arxiviq.substack.com dept.

https://arxiviq.substack.com/p/arc-is-a-vision-problem

( Read More... | no comments | on HN )

AlphaResearch: Accelerating New Algorithm Discovery with Language Models

Posted by che_shr_cat on Sunday November 16, 2025 @08:45PM
from the arxiviq.substack.com dept.

https://arxiviq.substack.com/p/alpharesearch-accelerating-new-algorithm

( Read More... | no comments | on HN )

LeJEPA: Provable and Scalable Self-Supervised Learning Without the Heuristics

Posted by che_shr_cat on Thursday November 13, 2025 @01:03PM
from the arxiviq.substack.com dept.

https://arxiviq.substack.com/p/lejepa-provable-and-scalable-self

( Read More... | 1 comment | on HN )

Nested Learning: The Illusion of Deep Learning Architectures

Posted by che_shr_cat on Monday November 10, 2025 @11:43AM
from the arxiviq.substack.com dept.

https://arxiviq.substack.com/p/nested-learning-the-illusion-of-deep

( Read More... | no comments | 2 points | on HN )

Context Engineering 2.0: The Context of Context Engineering

Posted by che_shr_cat on Thursday November 06, 2025 @05:57PM
from the arxiviq.substack.com dept.

https://arxiviq.substack.com/p/context-engineering-20-the-context

( Read More... | no comments | on HN )

A Practitioner's Guide to Kolmogorov-Arnold Networks

Posted by che_shr_cat on Sunday November 02, 2025 @12:56PM
from the arxiviq.substack.com dept.

https://arxiviq.substack.com/p/a-practitioners-guide-to-kolmogorov

( Read More... | no comments | on HN )

Kimi Linear: An Expressive, Efficient Attention Architecture

Posted by che_shr_cat on Saturday November 01, 2025 @06:53PM
from the arxiviq.substack.com dept.

https://arxiviq.substack.com/p/kimi-linear-an-expressive-efficient

( Read More... | no comments | on HN )

The Principles of Diffusion Models (470-pages)

Posted by che_shr_cat on Saturday November 01, 2025 @01:49PM
from the arxiviq.substack.com dept.

https://arxiviq.substack.com/p/the-principles-of-diffusion-models

( Read More... | no comments | on HN )

CaT Replaces CoT-SC / Compute as Teacher: Turning Inference Compute Into

Posted by che_shr_cat on Thursday October 23, 2025 @02:15PM
from the arxiviq.substack.com dept.

https://arxiviq.substack.com/p/compute-as-teacher-turning-inference

( Read More... | no comments | on HN )

Tiny Recursive Model (TRM) vs. Hierarchical Reasoning Model (HRM)

Posted by che_shr_cat on Sunday October 19, 2025 @02:41PM
from the gonzoml.substack.com dept.

https://gonzoml.substack.com/p/tiny-recursive-model

( Read More... | no comments | 2 points | on HN )

Barbarians at the Gate: How AI Is Upending Systems Research

Posted by che_shr_cat on Wednesday October 15, 2025 @09:33AM
from the arxiviq.substack.com dept.

https://arxiviq.substack.com/p/barbarians-at-the-gate-how-ai-is

( Read More... | no comments | 2 points | on HN )

Evolution Strategies at Scale: LLM Fine-Tuning Beyond Reinforcement Learning

Posted by che_shr_cat on Thursday October 09, 2025 @05:53AM
from the arxiviq.substack.com dept.

https://arxiviq.substack.com/p/evolution-strategies-at-scale-llm

( Read More... | no comments | on HN )

Autoreview: The Dragon Hatchling – The Missing Link Between the Transformer and

Posted by che_shr_cat on Sunday October 05, 2025 @05:55AM
from the arxiviq.substack.com dept.

https://arxiviq.substack.com/p/the-dragon-hatchling

( Read More... | no comments | on HN )

Stochastic Activations

Posted by che_shr_cat on Saturday October 04, 2025 @06:05PM
from the gonzoml.substack.com dept.

https://gonzoml.substack.com/p/stochastic-activations

( Read More... | no comments | on HN )

LLM-JEPA: Large Language Models Meet Joint Embedding Predictive Architectures

Posted by che_shr_cat on Monday September 22, 2025 @06:31AM
from the arxiviq.substack.com dept.

https://arxiviq.substack.com/p/llm-jepa-large-language-models-meet

( Read More... | no comments | on HN )

Review: SpikingBrain Technical Spiking Brain-Inspired Large Models

Posted by che_shr_cat on Saturday September 13, 2025 @08:14AM
from the arxiviq.substack.com dept.

https://arxiviq.substack.com/p/spikingbrain-technical-report-spiking

( Read More... | no comments | on HN )

K2-Think: A Parameter-Efficient Reasoning System

Posted by che_shr_cat on Friday September 12, 2025 @11:53AM
from the arxiviq.substack.com dept.

https://arxiviq.substack.com/p/k2-think-a-parameter-efficient-reasoning

( Read More... | 1 comment | 2 points | on HN )

Canaries in the Coal Mine? Six Facts about the Recent Employment Effects of AI

Posted by che_shr_cat on Sunday September 07, 2025 @07:35AM
from the arxiviq.substack.com dept.

https://arxiviq.substack.com/p/canaries-in-the-coal-mine-six-facts

( Read More... | no comments | on HN )

Fantastic Pretraining Optimizers and Where to Find Them

Posted by che_shr_cat on Wednesday September 03, 2025 @07:13AM
from the arxiviq.substack.com dept.

https://arxiviq.substack.com/p/fantastic-pretraining-optimizers

( Read More... | no comments | 2 points | on HN )

Solving the compute crisis with physics-based ASICs

Posted by che_shr_cat on Sunday August 31, 2025 @12:35PM
from the arxiviq.substack.com dept.

https://arxiviq.substack.com/p/solving-the-compute-crisis-with-physics

( Read More... | no comments | 4 points | on HN )

Critiques of World Models

Posted by che_shr_cat on Wednesday August 27, 2025 @05:38PM
from the arxiviq.substack.com dept.

https://arxiviq.substack.com/p/critiques-of-world-models

( Read More... | no comments | 2 points | on HN )

Re:Deep Think with Confidence on Sunday August 24, 2025 @12:50PM (#45005720)

I'm the author of this blog. That's correct, the texts are generated and then validated manually by me.

I also do manual reviews (https://gonzoml.substack.com/), but there are many more papers for which I don't have time to write a review. So I created a multi-agentic system to help me, and I'm constantly iterating to improve it. And I like the result. It was also validated by the paper authors a couple of times, they agree the reviews are correct. So, if you see something is definitely wrong, please let me know.

Regarding myself, I became at least x10 more productive in reading papers and understanding what's happening. Hope, it will also help some of you.

Deep Think with Confidence

Posted by che_shr_cat on Sunday August 24, 2025 @10:41AM
from the arxiviq.substack.com dept.

https://arxiviq.substack.com/p/deep-think-with-confidence

( Read More... | 35 comments | 63 points | on HN )

V-JEPA 2: Scaling V-JEPA

Posted by che_shr_cat on Friday August 22, 2025 @10:39AM
from the gonzoml.substack.com dept.

https://gonzoml.substack.com/p/v-jepa-2-scaling-v-jepa

( Read More... | no comments | 2 points | on HN )