II.
TestSet overview
Reference · livetest-set:tau-bench-airline-retail
tau-bench airline and retail domains overview
Source-backed tau-bench domain artifact for multi-turn tool-use and policy-following evaluation.
Attributes
displayName
tau-bench airline and retail domains
benchmarkId
releasedAt
2024-06-17
composition
Original tau-bench customer-service domains for tool-agent-user
interactions in airline and retail settings, with simulated users,
policy constraints, and domain APIs.
homepageUrl
description
Source-backed tau-bench domain artifact for multi-turn tool-use and
policy-following evaluation.
Outgoing edges
belongs_to_benchmark1
- benchmark:tau-bench·Benchmarktau-bench
Incoming edges
None.