AI-powered assistant for academic peer review. Built with LangGraph, this tool validates references against claims, flags unsupported assertions, performs literature reviews, and suggests relevant citations — helping reviewers and researchers assess rigor more efficiently.

The AI Scientist-v2: Workshop-Level Automated Scientific Discovery via Agentic Tree Search

aiseo-audit analyzes a URL and reports how effectively its content can be ingested, understood, and reused by generative AI engines like ChatGPT, Claude, Gemini, and Perplexity.

Fork of vllm with targeted support for mechinterp

Runpod but easy & fun

open-source JavaScript/TypeScript library that renders QTI 3 (Question and Test Interoperability) XML content in web applications. The library provides a framework-agnostic component that any developer can install via NPM and integrate into their front-end application to display standards-compliant assessment items.

An mcp server built for fathom and works in claude.

Supervised learning of SelfIE adapters on frozen language models - dataset prep, training, and all evals

Code to reproduce: Endogenous Steering Resistance in LLMs - spontaneous detection and correction of inappropriate activation steering

Extract hidden states from intermediate transformer layers using vLLM, with correct per-request tracking.