PyTOD: Programmable Task-Oriented Dialogue with Execution Feedback
By: Alexandru Coca , Bo-Hsiang Tseng , Pete Boothroyd and more
Potential Business Impact:
Helps computers understand conversations better.
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.
Similar Papers
Improving Multi-turn Task Completion in Task-Oriented Dialog Systems via Prompt Chaining and Fine-Grained Feedback
Computation and Language
Helps computers finish tasks by talking.
Planning with Diffusion Models for Target-Oriented Dialogue Systems
Computation and Language
Helps chatbots plan conversations better to reach goals.
Spec-TOD: A Specialized Instruction-Tuned LLM Framework for Efficient Task-Oriented Dialogue Systems
Computation and Language
Teaches computers to talk with less training.