Neural Amortized Inference for Nested Multi-Agent Reasoning

Authors: Kunal Jha, Tuan Anh Le, Chuanyang Jin, Yen-Ling Kuo, Joshua B. Tenenbaum, Tianmin Shu

AAAI 2024 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We evaluate our method in two challenging multi-agent interaction domains. The experimental results demonstrate that our method is computationally efficient while exhibiting minimal degradation in accuracy.
Researcher Affiliation Collaboration Kunal Jha1, Tuan Anh Le2, Chuanyang Jin3, Yen-Ling Kuo4, Joshua B. Tenenbaum5, Tianmin Shu5,6 1Dartmouth College 2Google Research 3New York University 4University of Virginia 5Massachusetts Institute of Technology 6Johns Hopkins University
Pseudocode No The paper describes algorithmic steps in paragraph format but does not include structured pseudocode or clearly labeled algorithm blocks.
Open Source Code Yes 1The code and the supplementary material are available at https: //www.tshu.io/nested reasoning.
Open Datasets No The paper describes synthesizing its own datasets for experiments ('This allows us to create two training sets S1... and S2...', 'create three training sets: S0...'), and mentions that 'We describe how we generate the training data in the supplementary material.' However, it does not provide concrete access information (like a link or citation to a public repository) for the generated datasets themselves, only the code to generate them.
Dataset Splits No The paper mentions 'training sets' and 'testing episodes' but does not explicitly provide details about validation dataset splits (e.g., percentages, sample counts, or methods for creating a validation set).
Hardware Specification No The paper does not provide any specific details about the hardware (e.g., GPU models, CPU types, memory) used to conduct the experiments.
Software Dependencies No The paper mentions using 'CARLO (Cao et al. 2020)' for the driving environment but does not specify version numbers for any software, libraries, or frameworks used in their implementation.
Experiment Setup Yes We evaluate our method in a 2D grid-world domain, Construction... The two agents and ten colored blocks are randomly spawned in 20 20 grid. There are 45 possible goals for Alice and two possible goals for Bob.