MRKL system
In plain wordsAn early design, pronounced 'miracle', in which a language model acts as a switchboard, sending each part of a question to the right specialist — a calculator, a database, a search tool — and combining the answers.
A modular, neuro-symbolic architecture ('Modular Reasoning, Knowledge and Language') in which a router built around a large language modellarge language model, n.: A computer program trained on enormous amounts of text to predict the words that come next. At vast scale, that simple skill lets it chat, answer questions, translate, summarize and write code. sends each input, or part of it, to one of several expert modules — neural models or symbolic components such as calculators, currency converters, databases and APIs — and composes their outputs into an answer. The motivation was that language models alone lack current knowledge, reliable arithmetic and access to proprietary data, and that retraining them is costly; delegating to discrete experts keeps the model general while making particular capabilities exact and updatable. AI21 Labs implemented the idea as Jurassic-X. The design anticipated tool usetool use, n.: An AI model's ability to reach beyond its own knowledge by operating outside tools — searching the web, running code, using a calculator or an app — and folding the results into its work. and function callingfunction calling, n.: A feature of AI model services that lets the model reply with a neatly formatted request to run one of a program's functions, rather than with ordinary text; the program, not the model, then runs it., and LangChain's first agents (late 2022) were described as MRKL agents that used ReActReAct, prop. n.: A method in which an AI alternates between thinking out loud about what to do next and taking an action, such as a web search, then uses what it observed to plan its next step.-style prompting to choose tools.
LangChain's original zero-shot agent was essentially a MRKL system with ReAct prompting.
Origin
Proposed by Ehud Karpas, Omri Abend, Yonatan Belinkov, Yoav Shoham, Amnon Shashua and colleagues at AI21 Labs in a paper posted on 1 May 2022; the acronym was chosen to be pronounced 'miracle'.
Related
- tool usetool use, n.: An AI model's ability to reach beyond its own knowledge by operating outside tools — searching the web, running code, using a calculator or an app — and folding the results into its work.
- ReActReAct, prop. n.: A method in which an AI alternates between thinking out loud about what to do next and taking an action, such as a web search, then uses what it observed to plan its next step.
- augmented language modelaugmented language model, n.: A language model hooked up to extras — a search engine, a calculator, a memory store or other tools — so it can do things a model working only from its training cannot.
- neuro-symbolic AIneuro-symbolic AI, n.: AI that combines neural networks, which learn patterns from data, with symbolic methods, which follow explicit rules and logic, aiming for systems that both learn flexibly and reason reliably.
- ToolformerToolformer, prop. n.: A 2023 Meta research model that learned on its own when a tool — a calendar, a translator, a search engine, a calculator — would help it predict text, keeping only the tool uses that actually paid off.
- agent frameworkagent framework, n.: A software library that gives developers ready-made parts for building AI agents — model connections, tool handling, memory, multi-agent coordination and logging — so they need not write everything from scratch.
Sources
- Karpas, E., Abend, O., Belinkov, Y. et al. (2022). MRKL Systems: A modular, neuro-symbolic architecture that combines large language models, external knowledge sources and discrete reasoning. arXiv:2205.00445. arxiv.org/abs/2205.00445
Mentioned in
AI agentAI agent, n.: An AI system that does not just answer but acts: given a goal, it decides on steps, uses tools such as a browser or code editor, checks the results and keeps going until the job is done., ReActReAct, prop. n.: A method in which an AI alternates between thinking out loud about what to do next and taking an action, such as a web search, then uses what it observed to plan its next step., tooltool, n.: Any outside capability an AI model is allowed to use — a web search, a calculator, a code runner, a database or an app — described to the model so it can ask for it by name., tool usetool use, n.: An AI model's ability to reach beyond its own knowledge by operating outside tools — searching the web, running code, using a calculator or an app — and folding the results into its work., augmented language modelaugmented language model, n.: A language model hooked up to extras — a search engine, a calculator, a memory store or other tools — so it can do things a model working only from its training cannot., ToolformerToolformer, prop. n.: A 2023 Meta research model that learned on its own when a tool — a calendar, a translator, a search engine, a calculator — would help it predict text, keeping only the tool uses that actually paid off.