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..
Multi-Agent Routing Value Iteration Network
Authors: Quinlan Sykora, Mengye Ren, Raquel Urtasun
ICML 2020 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We created a simulated environment to mimic realistic mapping performed by autonomous vehicles with unknown minimum edge coverage and traffic conditions; our approach significantly outperforms traditional solvers both in terms of total cost and runtime. We compare our approach with the following baselines. |
| Researcher Affiliation | Industry | Quinlan Sykora * Mengye Ren * Raquel Urtasun Correspondence to: Quinlan Sykora <EMAIL>, Mengye Ren <EMAIL>, Raquel Urtasun <EMAIL>. |
| Pseudocode | No | The paper describes the algorithm steps in paragraph form, but does not include a clearly labeled pseudocode or algorithm block. |
| Open Source Code | Yes | Our code and data are released at https://github.com/ uber/MARVIN |
| Open Datasets | Yes | Our code and data are released at https://github.com/ uber/MARVIN and The dataset contains 22,814 directed road graphs collected from 18 cities around the world. |
| Dataset Splits | Yes | We use a separate city for testing purposes and 10% of the training set for validation. |
| Hardware Specification | No | The paper does not provide specific hardware details such as GPU or CPU models used for running experiments. |
| Software Dependencies | No | The paper mentions using the Adam optimizer, but does not provide specific version numbers for software dependencies or libraries used in the implementation. |
| Experiment Setup | Yes | When training, we set the learning rate of our model to be 1e-3 using the Adam optimizer, with a decay rate of 0.1 every 2000 epochs. We train our model for 5000 epochs. We use a batch size of 50 graphs, each of which has up to 25 nodes. |