Condition-Level Interpretation of Task-Success Solution Spaces Using Probabilities of Causation

In the modelling and simulation of complex nonlinear and hybrid dynamical systems, forward simulation can generate task-success solution spaces even when success conditions cannot be derived as closed-form inequalities. However, interpreting the roles of individual parameter conditions becomes difficult as the dimensionality of the solution space increases. This study proposes Counterfactual Logical Dependency Analysis of Motor Task Success (CLDA-MTS), which provides a condition-level interpretation of task-success solution spaces generated by mechanistic dynamical models. CLDA-MTS compares the outcome of a candidate condition with those of reference alternatives under the same background condition and computes the probability of necessity, probability of sufficiency, and probability of necessity and sufficiency based on Pearl’s probabilities of causation (PoC). In the validation of CLDA-MTS using an inverted pendulum with first-order proportional control, for which analytical stability conditions are available, the success labels from forward simulation agreed with the analytical conditions at all sampled conditions, and the PoC values calculated from discrete samples were consistent with the continuous PoC limits and discretization errors. The positive, zero, and undefined classifications of PoC also agreed with those expected from the analytical solution. The Spring-Loaded Inverted Pendulum (SLIP) model was used to evaluate applicability and interpretive validity where closed-form stability conditions are difficult to derive. CLDA-MTS was applied to numerically detected periodic running solutions, and the resulting condition-level interpretations were logically consistent with established findings on stable touchdown-angle/leg-stiffness combinations. The geometry of the same detected-solution set provided a complementary representation of these results. Furthermore, changes in sufficiency from a single condition to a composite condition were quantified separately as components arising from changes in candidate outcomes and reference composition. These results demonstrate that CLDA-MTS provides a quantitative interpretation layer for extracting condition-level structure from simulation-generated task-success solution spaces.

Authors

Institutions

Publication Details

Journal
Modelling—International Open Access Journal of Modelling in Engineering Science
Published
2026-11-16
DOI
https://doi.org/10.3390/modelling7060256
Primary Topic
Simulation Techniques and Applications
Type
article
Field-Weighted Citation Impact
0.00

Funders

Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Condition-Level Interpretation of Task-Success Solution Spaces Using Probabilities of Causation

Takashi Kusaka, Shunsuke Nakamae, Akihiko Murai, Takayuki Tanaka
2 links
Modelling—International Open Access Journal of Modelling in Engineering Science
Simulation Techniques and Applications
article

Condition-Level Interpretation of Task-Success Solution Spaces Using Probabilities of Causation

Takashi Kusaka, Shunsuke Nakamae, Akihiko Murai, Takayuki Tanaka
article en

Abstract

In the modelling and simulation of complex nonlinear and hybrid dynamical systems, forward simulation can generate task-success solution spaces even when success conditions cannot be derived as closed-form inequalities. However, interpreting the roles of individual parameter conditions becomes difficult as the dimensionality of the solution space increases. This study proposes Counterfactual Logical Dependency Analysis of Motor Task Success (CLDA-MTS), which provides a condition-level interpretation of task-success solution spaces generated by mechanistic dynamical models. CLDA-MTS compares the outcome of a candidate condition with those of reference alternatives under the same background condition and computes the probability of necessity, probability of sufficiency, and probability of necessity and sufficiency based on Pearl’s probabilities of causation (PoC). In the validation of CLDA-MTS using an inverted pendulum with first-order proportional control, for which analytical stability conditions are available, the success labels from forward simulation agreed with the analytical conditions at all sampled conditions, and the PoC values calculated from discrete samples were consistent with the continuous PoC limits and discretization errors. The positive, zero, and undefined classifications of PoC also agreed with those expected from the analytical solution. The Spring-Loaded Inverted Pendulum (SLIP) model was used to evaluate applicability and interpretive validity where closed-form stability conditions are difficult to derive. CLDA-MTS was applied to numerically detected periodic running solutions, and the resulting condition-level interpretations were logically consistent with established findings on stable touchdown-angle/leg-stiffness combinations. The geometry of the same detected-solution set provided a complementary representation of these results. Furthermore, changes in sufficiency from a single condition to a composite condition were quantified separately as components arising from changes in candidate outcomes and reference composition. These results demonstrate that CLDA-MTS provides a quantitative interpretation layer for extracting condition-level structure from simulation-generated task-success solution spaces.

Modelling—International Open Access Journal of Modelling in Engineering ScienceVol. 7(6)
Hokkaido University (JP), AIST Kashiwa (JP), National Institute of Advanced Industrial Science and Technology (JP)
Japan Society for the Promotion of Science, Japan Science and Technology Agency
Simulation Techniques and Applications
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon