PyTOD: Programmable Task-Oriented Dialogue with Execution Feedback

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Coca, Alexandru, Tseng, Bo-Hsiang, Boothroyd, Pete, Cheng, Jianpeng, Gaynor, Mark, Zhang, Zhenxing, Stacey, Joe, Guigue, Tristan, Alonso, Héctor Martinez, Séaghdha, Diarmuid Ó, Johannsen, Anders
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915455025807360
author Coca, Alexandru
Tseng, Bo-Hsiang
Boothroyd, Pete
Cheng, Jianpeng
Gaynor, Mark
Zhang, Zhenxing
Stacey, Joe
Guigue, Tristan
Alonso, Héctor Martinez
Séaghdha, Diarmuid Ó
Johannsen, Anders
author_facet Coca, Alexandru
Tseng, Bo-Hsiang
Boothroyd, Pete
Cheng, Jianpeng
Gaynor, Mark
Zhang, Zhenxing
Stacey, Joe
Guigue, Tristan
Alonso, Héctor Martinez
Séaghdha, Diarmuid Ó
Johannsen, Anders
contents Programmable task-oriented dialogue (TOD) agents enable language models to follow structured dialogue policies, but their effectiveness hinges on accurate state tracking. We present PyTOD, an agent that generates executable code to track dialogue state and uses policy and execution feedback for efficient error correction. To this end, PyTOD employs a simple constrained decoding approach, using a language model instead of grammar rules to follow API schemata. This leads to state-of-the-art state tracking performance on the challenging SGD benchmark. Our experiments show that PyTOD surpasses strong baselines in both accuracy and robust user goal estimation as the dialogue progresses, demonstrating the effectiveness of execution-aware state tracking.
format Preprint
id arxiv_https___arxiv_org_abs_2508_15456
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle PyTOD: Programmable Task-Oriented Dialogue with Execution Feedback
Coca, Alexandru
Tseng, Bo-Hsiang
Boothroyd, Pete
Cheng, Jianpeng
Gaynor, Mark
Zhang, Zhenxing
Stacey, Joe
Guigue, Tristan
Alonso, Héctor Martinez
Séaghdha, Diarmuid Ó
Johannsen, Anders
Computation and Language
Programmable task-oriented dialogue (TOD) agents enable language models to follow structured dialogue policies, but their effectiveness hinges on accurate state tracking. We present PyTOD, an agent that generates executable code to track dialogue state and uses policy and execution feedback for efficient error correction. To this end, PyTOD employs a simple constrained decoding approach, using a language model instead of grammar rules to follow API schemata. This leads to state-of-the-art state tracking performance on the challenging SGD benchmark. Our experiments show that PyTOD surpasses strong baselines in both accuracy and robust user goal estimation as the dialogue progresses, demonstrating the effectiveness of execution-aware state tracking.
title PyTOD: Programmable Task-Oriented Dialogue with Execution Feedback
topic Computation and Language
url https://arxiv.org/abs/2508.15456