Tango: Declarative Semantics for Multiagent Communication Protocols
Authors: Munindar P. Singh, Samuel H. Christie V.
IJCAI 2021 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We show empirically that our algorithm produces smaller models and verifies correctness of sample protocols much faster than previous work. and Table 3 shows the total number of nodes (representing configurations or states), branches, and elapsed time for several sample protocols, including the examples in this paper. |
| Researcher Affiliation | Academia | 1Department of Computer Science, North Carolina State University, Raleigh, NC 27695, USA 2School of Computing and Communications, Lancaster University, Lancaster LA1 4WA, UK |
| Pseudocode | No | The paper describes rules and processes (e.g., Section 3.1, 3.2, 4.2) but does not present them in a formally labeled 'Pseudocode' or 'Algorithm' block. |
| Open Source Code | Yes | Our source code, examples, and instructions are available publicly in an online supplement at https://gitlab.com/masr/. |
| Open Datasets | No | The paper evaluates on 'sample protocols' such as 'PO Pay Cancel Ship' and 'Block Contra', which are protocol definitions, not publicly available datasets in the typical sense of a collection of data samples with a specific access method (link, DOI, citation). |
| Dataset Splits | No | The paper describes protocol verification and model generation but does not involve standard machine learning training/validation/test data splits or explicit dataset partitioning information. |
| Hardware Specification | Yes | All experiments were performed on the same Linux laptop, a Thinkpad T460 with an Intel i7-6600U CPU and 16GB of DDR3 RAM. |
| Software Dependencies | No | The paper states 'We have implemented Tango in Python,' but does not specify the version of Python or any other software dependencies with version numbers. |
| Experiment Setup | No | The paper mentions 'observations are sorted alphabetically by message name and nonsensitive observations are handled first' as part of the method, but does not provide specific experimental setup details such as hyperparameters, optimizer settings, or detailed training configurations. |