AnyTOD: A Programmable Task-Oriented Dialog System
This addresses the challenge of rapidly adapting dialog systems to new tasks for developers, reducing data and training needs, though it builds on existing neuro-symbolic approaches.
The authors tackled the problem of adapting task-oriented dialog systems to unseen tasks without task-specific training by proposing AnyTOD, a zero-shot system that treats dialog as a program executed by a language model, achieving state-of-the-art results on benchmarks like STAR, ABCD, and SGD.
We propose AnyTOD, an end-to-end, zero-shot task-oriented dialog (TOD) system capable of handling unseen tasks without task-specific training. We view TOD as a program executed by a language model (LM), where program logic and ontology is provided by a designer as a schema. To enable generalization to unseen schemas and programs without prior training, AnyTOD adopts a neuro-symbolic approach. A neural LM keeps track of events occurring during a conversation and a symbolic program implementing the dialog policy is executed to recommend next actions AnyTOD should take. This approach drastically reduces data annotation and model training requirements, addressing the enduring challenge of rapidly adapting a TOD system to unseen tasks and domains. We demonstrate state-of-the-art results on STAR, ABCD and SGD benchmarks. We also demonstrate strong zero-shot transfer ability in low-resource settings, such as zero-shot on MultiWOZ. In addition, we release STARv2, an updated version of the STAR dataset with richer annotations, for benchmarking zero-shot end-to-end TOD models.