English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Zendy Plus

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Zendy Tools

Zendy Open

Instruction-tuned large language models (LLMs) excel at many tasks but oftenfail to use external tools due to complicated and unfamiliar syntaxconstraints. While extensive fine-tuning and prompting can mitigate the issue,these approaches are expensive and hard to generalize. Furthermore, becausesyntax constraints are only learned implicitly during fine-tuning, models stillmake frequent syntax errors. Motivated by the fact that these constraints canbe better satisfied explicitly with constrained decoding, we propose TOOLDEC, adecoding algorithm using finite state machines to force LLMs to follow toolsyntax. Our experiments show that TOOLDEC eliminates all syntax errors,achieving significantly better performance on various base models andbenchmarks. More surprisingly, when applied to generalist out-of-the-box LLMssuch as Mistral-Instruct, TOOLDEC improves its accuracy in tool use from theinitial 0% to an impressive 52%, matching the performance of specializedfine-tuned models such as ToolLLM.

Don't Fine-Tune, Decode: Syntax Error-Free Tool Use via Constrained Decoding