AtariBench: Evaluating Spatiotemporal Reasoning of Vision-Language Models in Atari Gameplay
cs.CV, cs.AI
Submitted: 2026-11-05
Updated: 2026-11-05
Code: https://github.com/Farama-Foundation/Arcade-Learning-Environment
Project page: https://mignonjia.github.io/AtariBench
Terminology
Sources
- Video-Bench: A Comprehensive Benchmark and Toolkit for Evaluating Video-based Large Language Models
- VideoGameBench: Can Vision-Language Models complete popular video games?
- Factorio Learning Environment
- Pok'eChamp: an Expert-level Minimax Language Agent
- TextGames: Learning to Self-Play Text-Based Puzzle Games via Language Model Reasoning
- GameEval: Evaluating LLMs on Conversational Games
- Atari-GPT: Benchmarking Multimodal Large Language Models as Low-Level Policies in Atari Games
- Gymnasium: A Standard Interface for Reinforcement Learning Environments
- OpenAI Gym
- OASIS: On-Demand Hierarchical Event Memory for Streaming Video Reasoning
Related papers
- TEA: Text Encoder Alignment for Robust Concept Erasure in Text-to-Image Models
- Do Egocentric Video-Language Models Capture Both Hand- and Object-Centric Cues?
- Aura: Consistent Multi-Subject Video Generation via VLM-Grounded Semantic Alignment
- PVRF: All-in-one Adverse Weather Removal via Prior-modulated and Velocity-constrained Rectified Flow
- DLWM: Diverse Latent World Models for Efficient Multimodal Reasoning
- Out of Sight, Still in Mind: Token Compression for Omni-LLMs