GroupTravelBench is the first benchmark for multi-user multi-turn travel planning by LLM agents, showing even top models achieve plan validity below 12%.
Juhyun Oh, Eunsu Kim, and Alice Oh
4 Pith papers cite this work. Polarity classification is still indexing.
4
Pith papers citing it
representative citing papers
LLMs extract explicit constraints effectively but struggle with implicit open-world requirements, structural biases in plans, and ineffective self-correction during travel planning.
Plan-RewardBench is a trajectory-level preference benchmark that evaluates how well reward models distinguish preferred agent trajectories from hard distractors across safety refusal, tool handling, complex planning, and error recovery tasks.
TravelBench is a new benchmark with three subtasks and ten cached real-world tools to evaluate LLM agents on realistic multi-turn travel planning and capability boundaries.
citing papers explorer
-
GroupTravelBench: Benchmarking LLM Agents on Multi-Person Travel Planning
GroupTravelBench is the first benchmark for multi-user multi-turn travel planning by LLM agents, showing even top models achieve plan validity below 12%.
-
Revisiting the Travel Planning Capabilities of Large Language Models
LLMs extract explicit constraints effectively but struggle with implicit open-world requirements, structural biases in plans, and ineffective self-correction during travel planning.
-
Aligning Agents via Planning: A Benchmark for Trajectory-Level Reward Modeling
Plan-RewardBench is a trajectory-level preference benchmark that evaluates how well reward models distinguish preferred agent trajectories from hard distractors across safety refusal, tool handling, complex planning, and error recovery tasks.
-
Beyond Itinerary Planning-A Real-World Benchmark for Multi-Turn and Tool-Using Travel Tasks
TravelBench is a new benchmark with three subtasks and ten cached real-world tools to evaluate LLM agents on realistic multi-turn travel planning and capability boundaries.