RoboPapers: Evaluation & Benchmarks
How do we know when a robot actually works?
We look at the benchmarks, datasets, and real-world trials used to measure skills like manipulation and navigation. We at RoboPapers then ask whether strong lab results hold up across new tasks, environments, and robots.
Tap to Preview

Vision Language Models are In-Context Value Learners

  • Ep#4 Vision Language Models are In-Context Value Learners

  • Vision Language Models are In-Context Value Learners

    • Open Ecosystem for Embodied AI

  • Ep#69: MolmoSpaces, an Open Ecosystem for Embodied AI

    • ManiSkill3

  • Ep#26: ManiSkill3: GPU Parallelized Robotics Simulation and Rendering for Generalizable Embodied AI

    • Vision Language Models are In-Context Value Learners

  • Dyna-2: A 1-Million-Hour Scaling Law for World-Action Models

  • Ep#99: DYNA-2: A 1 Million Hour Scaling Law for World-Action Models