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 [1].
Infrastructure for AI Agents
Authors: Alan Chan, Kevin Wei, Sihao Huang, Nitarshan Rajkumar, Elija Perrier, Seth Lazar, Gillian K Hadfield, Markus Anderljung
TMLR 2025 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Theoretical | We identify three functions for agent infrastructure: 1) attributing actions, properties, and other information to specific agents, their users, or other actors; 2) shaping agents interactions; and 3) detecting and remedying harmful actions from agents. We provide an incomplete catalog of research directions for such functions. For each direction, we include analysis of use cases, infrastructure adoption, relationships to existing (internet) infrastructure, limitations, and open questions. |
| Researcher Affiliation | Academia | Kevin Wei Harvard Law School Sihao Huang University of Oxford Nitarshan Rajkumar University of Cambridge Elija Perrier Australian National University Gillian K. Hadfield Johns Hopkins University Markus Anderljung Centre for the Governance of AI |
| Pseudocode | No | The paper describes conceptual ideas and research directions. It does not contain any explicit pseudocode or algorithm blocks. |
| Open Source Code | No | The paper does not contain any explicit statements about releasing source code for the methodology described, nor does it provide links to any code repositories. |
| Open Datasets | No | The paper is a conceptual work that proposes agent infrastructure and does not conduct experiments requiring datasets. Therefore, it does not mention any datasets used in its own research or provide access information for them. |
| Dataset Splits | No | The paper is a conceptual work and does not perform experiments or analyze datasets, therefore, no dataset split information is provided. |
| Hardware Specification | No | The paper is a conceptual work and does not describe any experiments that would involve specific hardware. Therefore, no hardware specifications are provided. |
| Software Dependencies | No | The paper presents a conceptual framework for agent infrastructure and does not describe the implementation of any software that would require specific software dependencies with version numbers. |
| Experiment Setup | No | The paper is a theoretical and conceptual work and does not detail any experimental setup, hyperparameters, or training configurations. |