CrafText Benchmark: Advancing Instruction Following in Complex Multimodal Open-Ended World
This addresses the gap in benchmarking for multimodal instruction following in volatile settings, though it is incremental as it focuses on evaluation rather than new agent methods.
The authors tackled the problem of evaluating instruction-following agents in complex, dynamic environments by introducing the CrafText benchmark, which includes 3,924 instructions with 3,423 unique words across multiple task types to assess generalization and adaptability.
Following instructions in real-world conditions requires the ability to adapt to the world's volatility and entanglement: the environment is dynamic and unpredictable, instructions can be linguistically complex with diverse vocabulary, and the number of possible goals an agent may encounter is vast. Despite extensive research in this area, most studies are conducted in static environments with simple instructions and a limited vocabulary, making it difficult to assess agent performance in more diverse and challenging settings. To address this gap, we introduce CrafText, a benchmark for evaluating instruction following in a multimodal environment with diverse instructions and dynamic interactions. CrafText includes 3,924 instructions with 3,423 unique words, covering Localization, Conditional, Building, and Achievement tasks. Additionally, we propose an evaluation protocol that measures an agent's ability to generalize to novel instruction formulations and dynamically evolving task configurations, providing a rigorous test of both linguistic understanding and adaptive decision-making.