← back

Research

Deep dives, technical analysis, and system breakdowns

Pinned

🪨 Crystals of Code
LLM generation = crystal growth (Boltzmann proof)

LLM generation and crystal growth are mathematically identical, proven by Boltzmann distribution. 10-level distillation of 200+ papers across statistical mechanics, crystallography, information theory, and ML. Code binds to APIs like molecules bind to substrates.

→ Full research

🤖 Agent Frameworks
8 systems • 7 concepts • 13 patterns

Comprehensive analysis of agent systems: arizuko, Claude Code, Cline, Devin, OpenHands, Cursor, Langchain, Autogen. Architectural trade-offs in tool invocation, context management, state persistence, multi-agent coordination, and autonomous execution.

→ Full analysis

🔍 Taste
AST-based style fragment extraction

Designing simplicity & taste tools for LLMs. Scan codebase → extract structural patterns → output natural language fragments for prompts. Novel approach: automated detection of boring code patterns (guard clauses, flat control flow, function length) to nudge LLM code generation.

→ Full research

🧬 Biotech Software Engineering
Career guide for bio software engineers

Comprehensive guide for software engineers entering biotech. Covers cellular factories, grown materials, software stack for bio R&D. From DNA assembly automation to fermentation scale-up, bioreactor control systems, and computational biology tooling.

→ Full guide

AI & LLMs

🧠 Self-Improving Agents
700 papers • 19 research axes

Reinforcement learning for coding agents. From reward modeling to policy optimization, self-play, and continual learning. Covers RLHF, outcome-based rewards, test-driven RL, multi-agent debate, self-correction, and iterative refinement strategies.

→ Full research

🔒 Anthropic Mythos Preview
AI cybersecurity capabilities

Analysis of Anthropic's AI-driven vulnerability discovery & exploitation capabilities. Claude Mythos autonomously discovers zero-days in major OSes and browsers. Disclosed vulnerabilities: FreeBSD NFS RCE (CVE-2026-4747), OpenBSD TCP SACK DoS, FFmpeg H.264 OOB, Linux kernel privesc chains. Over 99% remain unpatched.

→ Full analysis

Infrastructure & Systems

🎯 Virtue: Esoteric Languages
Brainfuck, Befunge, Whitespace, Unlambda

Compilers, interpreters, and visualizations for esoteric programming languages. Brainfuck (300-byte compiler), Befunge-93 (2D stack-based), Whitespace (only spaces/tabs/newlines), Unlambda (pure functional combinator calculus). Interactive playgrounds with step debuggers and execution traces.

→ Explore languages

⚖️ PocketBase vs Arizuko
Architectural comparison

PocketBase (SQLite + Go monolith) vs arizuko (microservice agent gateway). Trade-offs in simplicity, scaling, and agent orchestration. PocketBase optimizes for single-developer simplicity with embedded DB. Arizuko optimizes for multi-agent coordination with separate processes and IPC.

→ Full comparison

🛡️ CVE-2026-4747
FreeBSD NFS RCE • 17 years old

Critical vulnerability disclosure: unauthenticated remote → full root access. Stack buffer overflow (304 bytes past 128-byte buffer). No stack canary, no KASLR. Exploit uses 6-packet ROP chain to append attacker SSH key. Full autonomous discovery and exploitation by Claude Mythos in several hours.

→ Full technical analysis

← krons.fiu.wtf