$τ^2$-Bench: Evaluating Conversational Agents in a Dual-Control Environment
Victor Barres, Dong, Honghua, Ray, Soham, Si, Xujie, Karthik Narasimhan
De quem é
- University of Torontoafiliação declarada pelos autores, segundo o OpenAlex
- Vector Instituteafiliação declarada pelos autores, segundo o OpenAlex
Outros papers das mesmas organizações
- AI models collapse when trained on recursively generated dataUniversity of Toronto · 2.565 citações
- Mastering diverse control tasks through world modelsUniversity of Toronto · 2.495 citações
- PixelSplat: 3D Gaussian Splats from Image Pairs for Scalable Generalizable 3D ReconstructionMassachusetts Institute of Technology · 966 citações
- ConceptGraphs: Open-Vocabulary 3D Scene Graphs for Perception and PlanningUnited States Army Combat Capabilities Development Command · 834 citações
- The AI Scientist-v2: Workshop-Level Automated Scientific Discovery via Agentic Tree SearchUniversity of British Columbia · 734 citações
- LMDrive: Closed-Loop End-to-End Driving with Large Language ModelsUniversity of Toronto · 587 citações