Articles by nkko
2

Mistral 3 Large uses the exact DeepSeek V3 architecture (reddit.com)

3

GitHub Actions CPU performance benchmarks (runs-on.com)

1

Building a Self-Training AI System with LLM Agents: SelfAI (github.com/xiaoxiao-woo)

76

Shai-Hulud compromised a dev machine and raided GitHub org access: a post-mortem (trigger.dev)

1

Scale Test-Time Compute with Parallel Coordinated Reasoning [pdf] (github.com/stepfun-ai)

1

An open-source energy and latency profiler for LLMs: ELANA (arxiv.org)

1

Fine-tuning Gemma 3 for mobile (googleblog.com)

2

Growth Is Value Flow, Not Vanity Metrics (nibzard.com)

1

How Prompt Caching Works (sankalp.bearblog.dev)

1

Devtools Just Became AI Infrastructure (nibzard.com)

1

Google Introducing Agent Sandbox (cloud.google.com)

5

Most of What We Call Progress (yusufaytas.com)

2

Aardvark: OpenAI's Agentic Security Researcher (openai.com)

2

Why do AI models use so many em-dashes? (seangoedecke.com)

2

Drug Discovery with Apex Protocol by Nvidia and Numerion Labs (arxiv.org)

1

The state of GitHub in 2025: AI leads TypeScript to #1 (github.blog)

1

Foxconn to deploy humanoid robots at Houston AI server plant (reuters.com)

1

Agent Labs Are Eating the Software World (nibzard.com)

3

Importing vs. Fetching JSON (jakearchibald.com)

1

Serving Humans and AI Through Content Negotiation (nibzard.com)

1

Quality software and the people who write it (vickiboykis.com)

1

Show HN: SlideGauge – Static analyzer for Marp Markdown decks (github.com/nibzard)

1

AI Agent Benchmark Compendium (philschmid.de)

3

Agents 2.0: From Shallow Loops to Deep Agents (philschmid.de)

2

DevRel is Back and we're not sure why (dx.tips)

2

How to Build an AI Startup That Lasts (nibzard.com)

1

Training a Deep Learning Model for Echogram Semantic Segmentation (oceanstream.io)

2

Googler Michel Devoret Awarded the Nobel Prize in Physics (blog.google)

3

Developer Trust over Conversion (nibzard.com)

2

Vercel vs. Cloudflare: two philosophies of building for developers (bharath.sh)

2

A universal API to cloud code execution sandboxes (cased.com)

2

Core Web Vitals (addyosmani.com)

1

Empirical Study of Pull Requests on GitHub (arxiv.org)

1

Graph-Based AI Compiler (tdcommons.org)

4

Building a Next-Generation Key-Value Store at Airbnb (medium.com/airbnb-engineering)

3

Development Productivity, Not Developer Productivity (redmonk.com)

1

Autonomous Workflow Agent Architecture (agentic-patterns.com)

1

2025 Dora State of AI-Assisted Software Development (cloud.google.com)

1

The Bottleneck in AI Development: Humans (nibzard.com)

10

TikTok's Ex-Algorithm Chief Launches Verdent AI Coding Tool (thenewstack.io)

2

Twitter's Open-Source Algorithm Analysis with Claude Code (nibzard.github.io)

1

How to harness skills that AI can't automate (figma.com)

1

Claude Code CLI with GLM-4.5 (nibzard.com)

1

Nano Banana in Java (glaforge.dev)

1

Survey: Most IT Teams Not Prepared to Manage AI Workloads (devops.com)

2

Kubernetes Primer: Dynamic Resource Allocation (DRA) for GPU Workloads (thenewstack.io)

1

AI Ate Its Own Tail, and I Learned Something About Writing (nibzard.com)

25

[flagged] Vibe Coding Through the Berghain Challenge (nibzard.com)

3

Ukraine's largest Starlink repair shop (technologyreview.com)

2

Show HN: Privacy first tone rewriter Chrome extension using local Gemini Nano (github.com/nibzard)

1

Show HN: Claude Threads – Browse Local Claude Code Sessions (github.com/nibzard)

1

Provider-agnostic, open-source evaluation infra for LLMs (github.com/groq)

1

I Built a Tool to Test AI's Command Line Ax (nibzard.com)

1

Coordination and Collaborative Reasoning in Multi-Agent LLMs (arxiv.org)

1

Why AI Dev Tools Need Different Growth (nibzard.com)

3

AI has killed free developer tools (redmonk.com)

1

Escalated, the AI Browser Wars Have – Quickly (spyglass.org)

2

Claude Code team joining Cursor (twitter.com/amir)

1

FAQ and Answers About AI Evals [pdf] (dropbox.com)

3

New MacBook with A18 Pro Chip Spotted in Apple Code (macrumors.com)

2

The future of labor in a world with increasingly productive AI (knightcolumbia.org)

1

Remote AI Runtime for Phoenix (Elixir) (phoenix.new)

1

TUI for Exploring LMDB Databases (github.com/nibzard)

1

Show HN: LLM Loop - An autonomous task execution plugin for the LLM CLI tool (github.com/nibzard)

2

Mary Meeker Artificial Intelligence Report [pdf] (bondcap.com)

3

Takeaways from Jony Ive's Stripe Sessions Chat (spyglass.org)

1

ChatGPT: H1 2025 Strategy (justice.gov)

3

Agentic Patterns (agentic-patterns.com)

2

AI Agents Are Reshaping Creation (nibzard.com)

2

Codestral Embed – embedding model for code (mistral.ai)

2

What Sourcegraph learned building AI coding agents (nibzard.com)

1

Claude Code Cheatsheet (nibzard.com)

2

Why Senior Engineers Overlook Small AI Wins (nibzard.com)

4

AI Coding Agent Pricing (nibzard.com)

2

Wine Variety Prediction with LLMs (github.com/ivanfioravanti)

2

Agent Building Is Software Engineering (matijagrcic.notion.site)

1

Tiny and Mighty Reward Models: J1 (github.com/haizelabs)

2

Evaluation of Engineering Artificial General Intelligence (arxiv.org)

1

AI Sandbox Benchmark (github.com/nkkko)

1

Superposition of Features Creates Power Law Performance in LLMs (arxiv.org)

2

LLM inference economics from first principles (tensoreconomics.com)

1

Intellect-2: A Reasoning Model Trained Through Globally Decentralized RL (arxiv.org)

4

Qwen3 Technical Report [pdf] (github.com/qwenlm)

1

Show HN: Hacker News Karma Tracker GitHub Action (github.com/nkkko)

169

LegoGPT: Generating Physically Stable and Buildable Lego (avalovelace1.github.io)

1

Model Soup's Recipe for Embeddings (jina.ai)

1

Base Models Beat Aligned Models at Randomness and Creativity (arxiv.org)

1

Seductive Language for Narcissists in Job Postings (informs.org)

2

Recursive Planning Based Open Writing Project: WriteHERE (github.com/principia-ai)

1

OSS Customer support agents by Antiwork (github.com/antiwork)

1

Real-time collaborative stream overlay (pogly.gg)

1

Evaluating Frontier Models for Stealth and Situational Awareness (arxiv.org)

1

Llama-Nemotron: Efficient Reasoning Models (arxiv.org)

63

Dummy's Guide to Modern LLM Sampling (rentry.co)

2

Physics of Language Models: Architecture Design and the Magic of Canon Layers (ssrn.com)

3

Groq's First Compound AI System (LLM with Compute) (groq.com)

2

Qwen3: Supervised Fine-Tuning with TRL (colab.research.google.com)

1

Redefining Agent Experience (daytona.io)

2

MCPs, Gatekeepers, and the Future of AI (iamcharliegraham.com)

1

Infrastructure for AI Agents (arxiv.org)