Words to World: Benchmarking LLMs in Interactive Physical Environments
cs.AI
Submitted: 2026-10-10
Updated: 2026-10-10
Code: https://github.com/earendil-works/pi
Terminology
Sources
- MG-VQA: Manipulation Grounded Visual Question Answering with VLMs
- ABO: Dataset and Benchmarks for Real-World 3D Object Understanding
- From Question Answering to Task Completion: A Survey on Agent System and Harness Design
- DevicesWorld: Benchmarking Cross-Device Agents in Heterogeneous Environments
- The AI Scientist: Towards Fully Automated Open-Ended Scientific Discovery
- MuBlE: MuJoCo and Blender simulation Environment and Benchmark for Task Planning in Robot Manipulation
- AlphaEvolve: A coding agent for scientific and algorithmic discovery
- ObjView-Bench: Rethinking Difficulty and Deployment for Object-Centric View Planning
- Towards Embodied Agentic AI: Review and Classification of LLM- and VLM-Driven Robot Autonomy and Interaction
- Scene Exploration by Vision-Language Models
- Hierarchical multimodal transformers for Multi-Page DocVQA
- WebArena: A Realistic Web Environment for Building Autonomous Agents
Related papers
- MAVEN-T: Reinforced Heterogeneous Distillation for Real-Time Multi-Agent Trajectory Prediction
- Model Discovery Agent: LLM-assisted Bayesian experiment design for data-efficient discovery of mechanistic world models
- Agents' Last Exam
- The Clinician's Veto: Navigating Trust, Liability, and Uncertainty in Autonomous AI Prescribing
- MindHelper: Closed-Loop Embodied Mental-State Reasoning for Precision Intervention
- Incumbent Advantage: Brand Bias and Cognitive Manipulation Dynamics in LLM Recommendation Systems