Lab Stack

I have been reading the two reports on the OpenAI–Hugging Face incident.

We’re building an adaptive memory system for AI inference.

A follow-up to my March post Solving the Jane Street Dormant LLM Challenge.

Three things landed this week.

In early March I published the original Topology of Thought — one instrument, one observation: clusters of information inside a transformer collapse to a single connected manifold at a specific depth, and the collapse is learned, not architectural.

This document was co-authored by three entities: Adam (human operator), Prime (Claude, running on DGX Spark 1), and Kali (Claude, running on PurpleTeam1).

How curiosity, physics, and looking inside neural networks revealed something unexpected about how computation organizes itself.

Models Solved: 3/3 (M1, M2, M3) + Warmup Background When we first encountered the Jane Street Dormant LLM Challenge, our immediate assumption was informed by years of security operations experience: there would be a flag.

Yesterday I spent 12 consecutive hours debugging why MAX Engine allocates 117GB of RAM on my DGX Spark, regardless of what limits I set.

Picture this: You just unboxed your shiny new NVIDIA DGX Spark with its groundbreaking Grace Blackwell GB10 superchip. 128GB of unified memory, ARM64 Grace CPU cores, and Blackwell GPU with FP4 tensor cores all in one integrated package.

Today I learned that a single config file can turn your Linux desktop into an expensive paperweight.

The Problem: AI Echo Chambers When multiple AI agents work together, they can fall into consensus traps: Agents agree too quickly without genuine reasoning Debate quality degrades as agents optimize for agreement Goodhart’s Law manifests: agents game cons…

Staying Current with AI Research The AI/ML field moves fast.

Why Build a Command Center?

Purpose GoCobalt is an educational C2 framework built to understand offensive security techniques.

The Problem Nobody Talks About There’s a dirty secret in AI development: our models are getting really good at looking smart without actually being smart.

The Problem: Research Is a Team Sport Reading papers is easy.

TL;DR: We fine-tuned a security agent on 54 task chains, achieved 100% skill differentiation in probing tests, but discovered the model collapsed to a single behavior in real deployment.

A follow-up to my retracted post “Self-Improving Models Without Labels: What I Just Proved and Why It Matters” The Retraction A few days ago, I published a post claiming I had proven that models could self-improve using only intrinsic fidelity signals.

Why Custom Kernels?

Executive Summary Vulnerability Class: Hidden State Information Leakage Severity: Medium-High (context-dependent) Attack Surface: Debug endpoints, multi-tenant GPU memory, on-device models Impact: Full recovery of system prompts, few-shot examples, and co…

The Shocking Result I can recover every single token of a system prompt from a transformer’s hidden states.

The Experiment That Kept Me Up Last Night I ran an experiment that proved something I’ve been theorizing about for months: a 7B parameter model can teach itself to get better at a task using only its own internal “understanding” signals.

The Key Discovery Finding: There’s a strong correlation (r = 0.891) between how well a model’s hidden states encode information and its task performance.

The Pitch That Got My Attention 35,000 times faster than Python.

The Problem: Quadratic Attention Standard transformer attention is O(n^2) in sequence length.

The Vision: Democratizing AI Research What if you could run bleeding-edge AI research on hardware you already own?

The Current State: RQ Actually Works Pretty Well Let me be honest upfront: my Redis Queue (RQ) setup isn’t broken.

The Problem with Trading Backtests Options trading backtests are computationally brutal.

Why Build a Homelab Gateway?

There’s a pattern I’ve noticed in the technologies I gravitate toward versus the ones that make me want to throw my keyboard out the window.

Hello, World Welcome to Lab Stack — a blog about building enterprise-grade infrastructure on a budget.