Responsibility Anticipation and Attribution in LTLf

Authors: Giuseppe De Giacomo, Emiliano Lorini, Timothy Parker, Gianmarco Parretti

IJCAI 2025 | Venue PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Theoretical In this paper, we study different variants of responsibility for LTLf outcomes based on strategic reasoning. We show a connection with notions in reactive synthesis, including the synthesis of winning, dominant, and best-effort strategies. This connection provides a strong computational grounding of responsibility, allowing us to characterize the worst-case computational complexity and devise sound, complete, and optimal algorithms for anticipating and attributing responsibility. We prove membership of checking active responsibility by exhibiting a sound and complete algorithm to solve it.
Researcher Affiliation Academia Giuseppe De Giacomo1,2 , Emiliano Lorini3 , Timothy Parker3 and Gianmarco Parretti2 1University of Oxford 2University of Rome La Sapienza 3IRIT, CNRS, Toulouse University, France EMAIL, EMAIL, EMAIL EMAIL
Pseudocode Yes We begin by giving an algorithm to check if a strategy σag is winning for φ under E, denoted CHECKWIN(φ, E, σag): 1. Construct the NFA N φ of φ, the DFA AE of E, and the DFA Aσag of σag; 2. Restrict AE to the environment winning region and obtain DFA A E; and 3. Check language nonemptiness of the product N = N φ A E Aσag.
Open Source Code No The paper does not contain any explicit statement about open-sourcing code, nor does it provide links to code repositories or mention code in supplementary materials.
Open Datasets No The paper is theoretical and does not conduct experiments on specific datasets. The examples provided (e.g., “The plant is watered”) are illustrative scenarios and do not refer to actual datasets.
Dataset Splits No The paper does not involve empirical experiments using datasets, therefore, no dataset splits are discussed.
Hardware Specification No The paper focuses on theoretical contributions, computational complexity, and algorithm design, without describing any experimental setup or hardware used for running experiments.
Software Dependencies No The paper describes theoretical algorithms and complexity analysis but does not mention specific software dependencies with version numbers used for implementing or executing experiments.
Experiment Setup No The paper is theoretical and focuses on formalizing concepts, analyzing complexity, and designing algorithms. It does not describe any empirical experimental setup, hyperparameters, or training configurations.