Mark Gravestock
A Practical Review of Mechanistic Interpretability for...
Initializing search
    • Home
    • TIL
    • Blog
    • Bookmarks
    • Tags
    • Home
    • TIL
        • Renovate allows dependency cooldowns
    • Blog
        • 2026
        • podcast
    • Bookmarks
        • JSON-LD Playground
        • Summary of the Amazon DynamoDB Service Disruption in Northern Virginia (US-EAST-1) Region | Chris Richardson
        • How to find, install, and manage MCP servers with the GitHub MCP Registry - The GitHub Blog
        • Mapping PostgreSQL Interval to Java Duration with Hibernate
        • The orchestration challenge of systems of intent represents one of the most critical obstacles to deploying AI agents in enterprise environments. 🔺 This challenge concerns how agents (systems that… | Anthony Alcaraz | 42 comments
        • Understanding Spec-Driven-Development: Kiro, spec-kit, and Tessl
        • vladmihalcea/hypersistence-utils: The Hypersistence Utils library (previously known as Hibernate Types) gives you Spring and Hibernate utilities that can help you get the most out of your data access layer.
        • What Kind of Programming is Natural Language Programming?
        • Enterprise-wide Software Architecture as DDD Living Documentation
        • The lethal trifecta for AI agents: private data, untrusted content, and external communication
        • How to map polymorphic JSON objects with JPA and Hibernate
        • AGENTS.md
        • Architect Elevator Blog
        • AsyncAPI Studio
        • Handling Events Coming in an Unknown Order
        • Measuring the Impact of Early-2025 AI on Experienced Open-Source Developer Productivity
        • Anthropic's Claude Sonnet 4.5: Memory Tool and Contextual Recall | Tony Seale posted on the topic | LinkedIn
        • Domain-Driven Transformation
        • The wall confronting large language models
        • Troubleshooting Java, Second Edition - Laurențiu Spilcă
        • UDA (Unified Data Architecture) is the foundation for connected data in Content Engineering at Netflix
        • Teaching large language models how to absorb new knowledge | MIT News | Massachusetts Institute of Technology
        • BugMagnet: A command for AI coding assistants to detect bugs and gaps | Gojko Adzic posted on the topic | LinkedIn
        • Slash commands - Claude Code Docs
        • A Practical Review of Mechanistic Interpretability for...
        • How to Make Claude Code Skills Activate Reliably
        • Mapping the Mind of a Large Language Model Anthropic
        • Scaling Monosemanticity: Extracting Interpretable Features from Claude 3 Sonnet
        • Understanding LLMs: Insights from Mechanistic Interpretability — LessWrong
        • My Foreword to "Frictionless"
        • Build Better Agents in Java vs Python: Embabel vs LangGraph
        • Requeuing Roulette in Event-Driven Architecture and Messaging
        • Spring Framework 7 and Spring Boot 4: The tastiest bites - JVM Weekly vol. 153
        • LLM APIs are a Synchronization Problem
        • Large Language Models’ Emergent Abilities Are a Mirage
        • CUPID: for joyful coding
        • Scaling Laws for Neural Language Models
        • Inside Azure Innovations with Mark Russinovich
        • OpenTelemetry with Spring Boot
        • Measuring the Impact of Early-2025 AI on Experienced Open-Source...
        • Consumers, projectors, reactors and all that messaging jazz in Emmett
        • Read-write and read-only transaction routing with Spring - Vlad Mihalcea
        • It’s The End Of Observability As We Know It (And I Feel Fine) | Honeycomb
        • Home | xRegistry Codegen
        • The Hierarchy Is Bullshit (And Bad For Business)
        • How to Build a Great Platform » Code Simplicity
        • Our 25 favourite European travel discoveries of 2025
        • Wednesday 17 December, 2025
        • Why “Microservices” Debates Miss the Point - CodeOpinion
        • How I wrote JustHTML using coding agents - Friendly Bit
        • 2025 LLM Year in Review
        • Prompt caching: 10x cheaper LLM tokens, but how? | ngrok blog
        • Open Data Mesh Platform
        • Embabel Year-End Update: Building The Best Agent Framework
        • Just Use Postgres for Everything
        • Open Social — overreacted
        • Resonant Computing Manifesto
        • https://arxiv.org/pdf/2512.15567
        • Redirect
        • Value driven technical decisions in software development
        • Authentication and authorization in a microservice architecture: Part 1 - Introduction
        • Fragments Dec 4
        • GitHub - muratcankoylan/Agent-Skills-for-Context-Engineering: A comprehensive collection of Agent Skills for context engineering, multi-agent architectures, and production agent systems. Use when building, optimizing, or debugging agent systems that require effective context management.
        • Research → Plan → Implement Pattern | goose
        • precision-medicine-mcp/architecture at main · lynnlangit/precision-medicine-mcp
        • Why Write Engineering Blogs
        • The best books of 2025
        • Who knows this source code?
        • [2512.21309] A Plan Reuse Mechanism for LLM-Driven Agent
        • A Complete 10-Day Japan Itinerary With Stops in Toyko, Kyoto and Osaka
        • Deploy a RAG chatbot app using Radius
        • E.W.Dijkstra Archive: On the foolishness of "natural language programming". (EWD 667)
        • How to Create a Test Data Builder | Code With Arho
        • Is DuckLake a Step Backward?
        • LMAX Disruptor: High performance alternative to bounded queues for exchanging data between concurrent threads
        • On Exactitude in Technical Debt
        • Strengthening the Inner Developer Loop: Turn AI Into a Reliable Engineering Partner
        • The LMAX Architecture
        • Turning Assertions Into a Domain-Specific Language - Petri Kainulainen
        • What Is A “Good Software Developer”, Anyway?
        • You Will Never Know Enough
        • Gunnar Morling
        • How Good Is AI at Coding React (Really)?
        • Architecturally Evident Applications– How to Bridge the Model-Code Gap
        • Rebuilding Event-Driven Read Models in a safe and resilient way
        • Coupling from a big-O perspective
        • A Software Library with No Code
        • Karabiner-Elements complex_modifications rules
        • Clean Contexts
        • gunnarmorling/streaming-examples: Example projects and demos around data streaming , stream processing, change data capture, and more.
        • Aspire for JavaScript developers
        • On rebuilding read models, Dead-Letter Queues and Why Letting Go is Sometimes the Answer
        • Singing the gospel of collective efficacy (Interconnected)
        • Software Survival 3.0. I spent a lot of time writing software… by Steve Yegge
        • Don’t believe the hype: AGI is far from inevitable | Radboud University
        • Software Development vs Maintenance: The True Cost Equation | Idea Link
        • The AI-Ready Software Developer: Conclusion – Same Game, Different Dice
        • Facts and Myths about CQRS • Oskar Dudycz • Devoxx Poland 2023
        • Fundamentals :: Spring Modulith
        • My Top 50 London Restaurants (2026 Edition)
        • CodeSpeak
        • Building production-ready Event-Sourced Node Apps with Ralph-Loop and AI
        • DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning
        • AI Doesn't Fix Your Real Bottleneck
        • Claude Code's Backlog Crisis: 6,383 Open Issues
        • Java on Kubernetes 2026: Fix Defaults, Boost Performance
        • Nick Tune - Portable Skills Configuration
        • Combine Testcontainers and Spring Boot with multiple containers
        • Hook-driven dev workflows with Claude Code - Nick Tune
        • Scaling PostgreSQL to power 800 million ChatGPT users
        • Simon Wardley on 'Fewer Engineers'
        • Humans and Agents in Software Engineering Loops
        • Claude Code workflows as DSL-driven domain models - Nick Tune
        • From SOLID to CUPID: Design Principles That Survive Production in Cloud-Native .NET
        • NTCoding/autonomous-claude-agent-team
        • How I Audit a Legacy Rails Codebase in the First Week
        • HTMX + Spring Boot - Awesome Java UI
        • Quoting Joseph Weizenbaum
        • JetBrains names the debt AI agents leave behind - The New Stack
        • Auto mode for Claude Code
        • 𝗟𝗟𝗠𝘀 𝗔𝗿𝗲 𝗡𝗼𝘁 𝗥𝗲𝗮𝗱𝗶𝗻𝗴 𝗬𝗼𝘂𝗿 𝗖𝗼𝗱𝗲
        • Allium LLM-native spec language
        • GitHub - jasisz/aver: Aver is a programming language for auditable AI-written code: verify in source, deploy with Rust, prove with Lean/Dafny · GitHub
        • java.evolved Code Snippets | java.evolved
        • Programming languages for AI
        • Are SOLID Principles Still Relevant in the AI Era?
        • everything-claude-code: The agent harness performance optimization system.
        • The one where Oskar explains Example Mapping
        • Mr. Chatterbox, or, The Modern Prometheus
        • Quoting Soohoon Choi
        • Harness engineering for coding agent users
        • What Comes After Systems Thinking
        • Will AI force code to evolve or make it extinct? - The New Stack
        • Claude Code's source code appears to have leaked: here's what we know
        • How A.I. Helped One Man (and His Brother) Build a $1.8 Billion Company
        • How Does Naming Affect Language Models on Code Analysis Tasks
        • Quoting Greg Kroah-Hartman
        • After Systems Thinking, Comes Ashby’s Law
        • Eight years of wanting, three months of building with AI - Lalit Maganti
        • Coding Is Dead, Long Live Programming — Ian Cooper - Staccato Signals
        • JUXT Blog: A bug on the dark side of the Moon
        • The Programmers's Guide to Co-Designing with Agents - david whitney, agile & software practices in london
        • tropes.md - Tropes
        • The Logic of Complex Systems
        • GitHub - ChrisTitusTech/winutil: Chris Titus Tech's Windows Utility - Install Programs, Tweaks, Fixes, and Updates · GitHub
        • The Friction That Stopped Waste — Ian Cooper - Staccato Signals
        • Feynman — The open source AI research agent
        • webshop-demo/.claude/skills/analyze-flaky-coverage/SKILL.md at main · flounder4130/webshop-demo · GitHub
        • Agentic Discovery of Neural Architectures: AIRA-Compose and AIRA-Design
        • Code Remix Summit: Postcard from Miami - JVM Weekly vol. 175
        • embabel/embabel-air: Airline chatbot demo for Embabel
        • Performance: 53% faster parse+render, 61% fewer allocations by tobi · Pull Request #2056 · Shopify/liquid · GitHub
        • Maintainability sensors for coding agents
        • Property-Based Testing with Claude red.anthropic.com
        • The AI Productivity Paradox Research Report
    • Tags
    ai llm

    A Practical Review of Mechanistic Interpretability for...

    A Practical Review of Mechanistic Interpretability for...

    arxiv.org · 17 Nov 2025

    Made with Material for MkDocs