digitado

About digitado

https://www.digitado.com.br

Posts by :

When Chaos Wins: noisy net eval with noise off gave wildly inconsistent results. Turning it back on fixed everything.

⋅

Running a Rainbow DQN ablation on Snake (C51 + dueling + noisy nets). When I evaluated checkpoints with noise off (mean weights, sigma zeroed out, the standard approach), the scores were all over the place. Some checkpoints averaged 78, others averaged 18. Training curve at those same points was perfectly stable. First instinct was a bug. Checked everything. It wasn’t. The worst case was at ep450K. Deterministic eval produced a bimodal distribution: ~25% of episodes scored near zero, […]

Ver mais

Liked Liked

E-Estate 1 Year Live: Washington DC Summit as Real Estate Tokenization Enters Its Next Phase

⋅

New York, USA, May 15th, 2026/Chainwire/–E Estate Group Inc. announced that it will host E-Estate 1 Year Live: Washington DC Summit on June 13, 2026, bringing together company leadership, agents, buyers, strategic partners, and guests interested in the future of blockchain-based real estate ownership. The summit will take place at The Watergate Hotel in Washington, D.C. and will mark one year since the launch of the E-Estate platform. The event is designed as a milestone gathering for the […]

Ver mais

Liked Liked

How should I plan my learning path for reinforcement learning courses?

⋅

Hi everyone, I have a question about planning my reinforcement learning studies. I’m currently a sophomore majoring in a non-CS field. My math background includes calculus, probability and statistics, linear algebra, and some mathematical analysis. I want to start learning reinforcement learning, but according to many recommendations, it seems I may also need additional math courses such as ODEs, real analysis, stochastic processes, etc. Is that really necessary at my current stage? Or would it be better to […]

Ver mais

Liked Liked

177 Blog Posts To Learn About Advertising

⋅

Let’s learn about Advertising via these 177 free blog posts. They are ordered by HackerNoon reader engagement data. Visit the Learn Repo or LearnRepo.com to find the most read blog posts about any technology. Good Advertising is dialogue with people … Lee Clow 1. This Is How Google Will Collapse Reporting on Google’s future with today’s facts 2. Nikkei Advertisement for Astar Network Receives the Backing of Record 329 Japanese Corporations Astar Network has broken a world record by […]

Ver mais

Liked Liked
How to Fine-Tune an LLM: SFT, LoRA, QLoRA and DPO Explained

How to Fine-Tune an LLM: SFT, LoRA, QLoRA and DPO Explained

⋅

This blog post discusses the details of what finetuning is, why it’s needed, and how we can finetune an LLM model with practical examples. The fine-tuning is what brings life to the LLM model. It’s a technique to make models adapt to a specific task, such as coding, writing poems or songs, classifying objects in an image, etc. A typical lifecycle of LLM training is depicted below. In the pretraining model, it only learns to predict the next token, […]

Ver mais

Liked Liked
L2 Distance was Giving Me Wrong Answers. Here’s the Metric That Fixed it.

L2 Distance was Giving Me Wrong Answers. Here’s the Metric That Fixed it.

⋅

L2 distance told me two audio fingerprints were similar. They weren’t. The problem wasn’t my model. It was the metric. If you have not read my last blog on vector index, it will help you understand why I am doing what I am doing. Access this story for free here The Problem With L2 OmniPulse revolves around WST fingerprints. They aren’t a single point in space. They are the distribution of energy across wavelet scales and time. Two fingerprints can have similar […]

Ver mais

Liked Liked
Nous Research Proposes Lighthouse Attention: A Training-Only Selection-Based Hierarchical Attention That Delivers 1.4–1.7× Pretraining Speedup at Long Context

Nous Research Proposes Lighthouse Attention: A Training-Only Selection-Based Hierarchical Attention That Delivers 1.4–1.7× Pretraining Speedup at Long Context

⋅

Training large language models on long sequences has a well-known problem: attention is expensive. The scaled dot-product attention (SDPA) at the core of every transformer scales quadratically Θ(N²) in both compute and memory with sequence length N. FlashAttention addressed this through IO-aware tiling that avoids materializing the full N×N attention matrix in high-bandwidth memory, reducing the memory footprint significantly, but the underlying Θ(N²) compute scaling remains. Researchers at Nous Research have introduced a new method called Lighthouse Attention […]

Ver mais

Liked Liked
I Built an AI Pilot That Plans Like a Robot and Dodges Like a Human

I Built an AI Pilot That Plans Like a Robot and Dodges Like a Human

⋅

A hybrid A* + Deep Reinforcement Learning system in Unity, an SR-71, 5 million training steps, and the architectural insight that made it work. When a fighter pilot flies a routine sortie, they don’t improvise — they follow the plan. When a missile lock alarm goes off, they don’t follow the plan — they improvise. Classical AI is great at the first thing. Reinforcement learning is great at the second. So why does almost everyone try to use one of them for both? That […]

Ver mais

Liked Liked

Warelay -> OpenClaw

⋅

In preparation for a lightning talk I’m giving at PyCon US this afternoon I decided to figure out how many names OpenClaw has actually had since that first commit back in November. Thanks to this first_line_history.py tool (code here) the answer, according to the Git history of the OpenClaw README, is: Warelay → CLAWDIS → CLAWDBOT → Clawdbot → Moltbot →🦞 OpenClaw Or in detail (the output from the tool): 2025-11-24T11:23:15+01:00 16dfc1a # Warelay — WhatsApp Relay CLI […]

Ver mais

Liked Liked
Why Production AI Agents Fail in Ways You Won’t See Coming (Part 1)

Why Production AI Agents Fail in Ways You Won’t See Coming (Part 1)

⋅

My practical fixes for costly blind spots It was 11:47 PM on a Tuesday when Marcus, a senior engineer I used to work with, dropped me a Slack message. His company’s finance team had just asked him: “Can you explain this AWS/OpenAI charge? $48,200. This month.” The agent had been live for three weeks. It passed all their staging tests. Ninety‑two percent on their internal eval suite. The team had high‑fived the launch. Then Marcus opened the logs and started scrolling. […]

Ver mais

Liked Liked

Load More

↑

Digitado © 2026