Notice: The reproducibility variables underlying each score are classified using an automated LLM-based pipeline, validated against a manually labeled dataset. LLM-based classification introduces uncertainty and potential bias; scores should be interpreted as estimates. Full accuracy metrics and methodology are described in Coakley et alK. L. Coakley, T. Snelleman, H. Hoos, and O. E. Gundersen, "The embrace of open science: An analysis of a decade of AI research and 56 800 conference papers," Under Review, 2026..
LaCAM: Search-Based Algorithm for Quick Multi-Agent Pathfinding
Authors: Keisuke Okumura
AAAI 2023 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Our exhaustive experiments reveal that La CAM is comparable to or outperforms state-of-the-art sub-optimal MAPF algorithms in a variety of scenarios, regarding success rate, planning time, and solution quality of sum-of-costs. |
| Researcher Affiliation | Academia | Keisuke Okumura Tokyo Institute of Technology Tokyo, Japan EMAIL |
| Pseudocode | Yes | Algorithm 1 shows an example implementation of La CAM. |
| Open Source Code | Yes | La CAM was coded in C++, available in the online supplementary material. |
| Open Datasets | Yes | Next, we tested La CAM using the MAPF benchmark (Stern et al. 2019), which includes a set of four-connected grids and start goal pairs for agents. |
| Dataset Splits | No | The paper mentions using '25 random scenarios' from the MAPF benchmark but does not specify a distinct validation split or any explicit train/validation/test splits for reproducibility. |
| Hardware Specification | Yes | The experiments were run on a desktop PC with Intel Core i9-7960X 2.8 GHz CPU and 64 GB RAM. |
| Software Dependencies | No | The paper states 'La CAM was coded in C++' and mentions using 'implementations coded by their respective authors' for baselines, but it does not provide specific version numbers for any libraries, frameworks, or compilers used. |
| Experiment Setup | Yes | The runtime limit was set to 30 s following (Stern et al. 2019). La CAM was run five times for each setting. The runtime limit was set to 1000 s. ... The agent order of the initial high-level node was in descending order of the distance between the start and goal... In other high-level nodes... we prioritized agents who are not on their goal... |