Field Notes · Opening

RAG-Safety-Bench: Reliable Evaluation of Retrieval-Augmented LLM Safety

RAG-Safety-Bench is a benchmark for evaluating how retrieval-augmented generation systems handle safety risks when retrieving from document sources and responding to requests for harmful content.

OpenAI Agents API

The OpenAI Agents API is a product offering from OpenAI that enables developers to build and deploy agent systems.

LLM Visualizer – Build a Transformer from Scratch

LLM Visualizer is a tool or resource for understanding transformer architecture by building one from scratch.