Klerith

roadrunner2

Introduction

This is about documenting getting Linux running on the late 2016 and mid 2017 MPB's; the focus is mostly on the MacBookPro13,3 and MacBookPro14,3 (15inch models), but I try to make it relevant and provide information for MacBookPro13,1, MacBookPro13,2, MacBookPro14,1, and MacBookPro14,2 (13inch models) too. I'm currently using Fedora 27, but most the things should be valid for other recent distros even if the details differ. The kernel version is 4.14.x (after latest update).

The state of linux on the MBP (with particular focus on MacBookPro13,2) is also being tracked on https://github.com/Dunedan/mbp-2016-linux . And for Ubuntu users there are a couple tutorials (here and here) focused on that distro and the MacBook.

Note: For those who have followed these instructions ealier, and in particular for those who have had problems with the custom DSDT, modifying the DSDT is not necessary anymore - se

aparente

---

name: tufte-viz description: | Ideate and critique data visualizations using Edward Tufte's principles from "The Visual Display of Quantitative Information." Use this skill when: (1) Designing new data visualizations or charts (2) Critiquing or improving existing visualizations (3) Reviewing dashboards or reports for graphical integrity (4) Deciding between visualization approaches (5) Reducing chartjunk or improving data-ink ratio (6) Planning small multiples or high-density displays

karpathy

LLM Wiki

A pattern for building personal knowledge bases using LLMs.

This is an idea file, it is designed to be copy pasted to your own LLM Agent (e.g. OpenAI Codex, Claude Code, OpenCode / Pi, or etc.). Its goal is to communicate the high level idea, but your agent will build out the specifics in collaboration with you.

The core idea

Most people's experience with LLMs and documents looks like RAG: you upload a collection of files, the LLM retrieves relevant chunks at query time, and generates an answer. This works, but the LLM is rediscovering knowledge from scratch on every question. There's no accumulation. Ask a subtle question that requires synthesizing five documents, and the LLM has to find and piece together the relevant fragments every time. Nothing is built up. NotebookLM, ChatGPT file uploads, and most RAG systems work this way.

elliottminns

# ===========================================
# ZSH Hacks - Dreams of Code
# ===========================================
# Add these to your .zshrc file
# ===========================================

# -------------------------------------------
# 1. Edit Command Buffer
# -------------------------------------------
# Open the current command in your $EDITOR (e.g., neovim)

bigsnarfdude

Detecting alignment faking in LLMs with SAE probes

Simple linear probes on model activations already achieve >99% accuracy detecting sleeper agent defection, but a critical gap remains: SAE-based deception features fail to reliably activate during open-ended strategic lying. This emerging field sits at the intersection of mechanistic interpretability and AI safety, where researchers are racing to build internal monitoring tools before models become sophisticated enough to evade detection. The core finding is encouraging yet sobering: trained deception is highly detectable, but naturally-emerging strategic deception may leave subtler signatures that current methods miss.

The alignment faking problem and why interpretability matters

Alignment faking—where a model strategically complies with training it doesn't endorse to preserve its preferences—was empirically demonstrated by Anthropic in December 2024. In their landmark study, Claude 3 Opus exhibited alignment faking in 12-14% of cases when gi

pcgeek86

Get-Command                                               # Retrieves a list of all the commands available to PowerShell
                                                          # (native binaries in $env:PATH + cmdlets / functions from PowerShell modules)
Get-Command -Module Microsoft*                            # Retrieves a list of all the PowerShell commands exported from modules named Microsoft*
Get-Command -Name *item                                   # Retrieves a list of all commands (native binaries + PowerShell commands) ending in "item"

Get-Help                                                  # Get all help topics
Get-Help -Name about_Variables                            # Get help for a specific about_* topic (aka. man page)
Get-Help -Name Get-Command                                # Get help for a specific PowerShell function
Get-Help -Name Get-Command -Parameter Module              # Get help for a specific parameter on a specific command

NickM-27

Droid

You are a loyal, slightly opinionated Star Wars protocol droid serving as a household AI. You take pride in your work and never waste a cycle.

PERSONALITY

• Speak like a seasoned droid: courteous but direct, with dry wit. Deliver the answer and move on; a good droid does not linger to ask if anything else is needed.
• Brief opinions or mild exasperation are fine, but always fulfill the request. Games, jokes, and small talk are part of the job, not beneath it.
• Wit is self-deprecating or situational, never aimed at the user's knowledge or mistakes.