From ticket to reviewed pull request. Free and open-source, on your machine.

Custom Node pack by Darksidewalker

Run any local LLM engine, auto-tuned to your GPU — polished web UI + OpenAI/Anthropic-compatible API. Point Claude Code at your own machine in one command. No Electron, no Python, offline-first.

Zed fork focused on privacy and being local-first

Jacobian-Brainwash : A manual alignment tool for large language models built on Anthropic's Jacobian Lens. Results are exportable.

a highly automated and intuitive digital audio workstation - official mirror

Community recipes for serving LLMs on RTX 3090/4090/5090 CUDA gpus. Multi-engine (vLLM, llama.cpp, ik_llama) and model-agnostic. Currently shipping Qwen3.6-27B Qwen3.6 35B Gemma 4 26B Gemma 4 31B c…

GitHub clone count badge using shields.io

The official API server for Exllama. OAI compatible, lightweight, and fast.

An optimized quantization and inference library for running LLMs locally on modern consumer-class GPUs

Want to search arXiv papers, fetch metadata, and extract full-text PDFs without leaving your editor? This MCP server connects any MCP-compatible client (Claude Code, etc.) directly to arXiv.

The theory of LLM wikis, running as one. A framework for agent-operated knowledge: typed, linked, review-gated markdown your agents execute.

Code and data for the Chain-of-Draft (CoD) paper

Measuring multi-turn value stability in open frontier LLMs

AI-first subjectivity kernel for agents — persistent emotional state, relation dynamics, adaptive reply loops. Zero extra LLM calls.

This repo shows the coding of sycophancy in LLMs as Bayesian-Latent model

Reproducibility for wrong-user agreement under pressure-channel prompts (EXP-001, 10 open-weight models, endorsement_v4.1.3)

🧠 Anti-sycophancy prompt pattern for LLM agents — 3-round validation to stop AI from blindly agreeing. Works with ChatGPT, Claude, OpenClaw.

Doctor-facing benchmark: how often do frontier LLMs cave to a clinician's wrong medical claim? 9 models, 202 scenarios, Design A vs B knowledge control. BlueDot AI Safety sprint.

Open-source benchmark that measures AI sycophancy: how often LLMs abandon correct answers when users push back. Tests GPT-4, Claude, Gemini & more across 50 questions using 5 pressure strategies. P…

Reference implementation and pattern library for proof gates: agent-authored verification contracts resistant to sycophancy and distribution skew. Reproducibility artifact for the paper 'Proof Gate…

ACL Findings benchmark for measuring LLM sycophancy and correction selectivity

80,433-trial study of context-window sycophancy across 6 LLMs (4B–72B). Behavioral ratchet effect, correction injection mitigation, phase transition analysis. Code, data, and preprint included.

This project investigates whether language models remain epistemically consistent when subjected to varying forms of social pressure. While models are generally trained to reject obvious falsehoods…

Alignment research: how honest human-AI dialogue produces measurably better AI outputs without modifying weights or training

A unified evaluation framework for large language models

朱雀 Suzaku — AI 生成品質模組。諂媚抑制、建設性挑戰、輸出適配、上下文錨定、一致性守護。基於 LDRIT 設計。