Toolformer lets language models teach themselves to use external tools

Meta researchers introduce Toolformer, a language model training method that enables self-supervised learning of when and how to call external tools via simple APIs, such as calculators and search engines, without sacrificing core language abilities.

Large language models display strong few-shot and instruction-following abilities but still perform poorly on basic operations like arithmetic and factual lookup, where smaller specialized systems excel. Toolformer addresses this gap by enabling a language model to learn how to call external tools through simple APIs, combining general-purpose reasoning with the precision of dedicated components. The approach focuses on making the model decide autonomously which tools to use and how to integrate their outputs while preserving its underlying language modeling capabilities.

Toolformer is trained to determine which APIs to call, when to call them, what arguments to provide, and how to feed the returned results back into subsequent token prediction. Training proceeds in a self-supervised manner with only a handful of demonstrations required for each API, avoiding the need for large bespoke labeled datasets. The model effectively learns a policy over tool usage as part of standard next-token prediction, so tool calls become a natural extension of its text generation process instead of a separate control mechanism.

The researchers incorporate a diverse set of tools into Toolformer, including a calculator, a question and answer system, a search engine, a translation system, and a calendar. Toolformer achieves substantially improved zero-shot performance across a variety of downstream tasks and is often competitive with much larger models, while maintaining its core language modeling performance. The work suggests that carefully integrating external tools via self-supervised learning can allow language models to overcome persistent weaknesses in areas such as computation and factual retrieval, without scaling model size or sacrificing fluency.

65

Impact Score

Intel shuts down software defined silicon paywall for server features

Intel has quietly ended its software defined silicon On Demand program for Xeon servers after customers rejected the idea of paying extra to unlock built-in hardware features. The move signals a pullback from hardware paywalls that had raised concerns about feature gating beyond traditional software subscriptions.

Discord rolls out global age verification and teen default settings

Discord is introducing global teen-by-default settings in early March 2026, requiring age verification via government ID or facial scan to access age-gated content. The rollout expands an existing system used in the UK and Australia and is already drawing privacy concerns.

Contact Us

Got questions? Use the form to contact us.

Contact Form

Clicking next sends a verification code to your email. After verifying, you can enter your message.