Selected work

Representative Papers

From Optimal Policies to Individual Differences: Rethinking Reinforcement Learning for Biology

Jul 17, 2026

This study addresses a key limitation in existing reinforcement learning approaches, which often overlook individual differences when modeling biological behavior, focusing instead on optimal policies or population averages. To overcome this constraint, the work introduces a biologically interpretable framework that integrates methods from multiple subfields of reinforcement learning to construct a computational model capable of generating diverse individual behaviors. By systematically synthesizing technical strategies that support behavioral diversity, the research establishes a novel paradigm for modeling individual variation in biological agents. This paradigm effectively narrows the gap between simulated and real-world biological behaviors, offering both a theoretical foundation and practical guidance for future research in biologically plausible behavior modeling.

0 citationsRead paper

Social-spatial dependencies for learning visual navigation

Jul 08, 2026

This study addresses the limitation of traditional visual navigation approaches, which often neglect the influence of social structures and spatial interactions on individual behavioral strategies within groups. The authors propose a bottom-up modeling framework that integrates deep reinforcement learning with multi-agent simulation to train agents in non-stationary dynamic environments, thereby investigating how task objectives and socio-spatial factors jointly shape the evolution of navigation strategies. Their findings demonstrate that high-quality social information can induce strategic phase transitions, successfully reproducing behavioral shifts ranging from individual navigation and group following to crowd-aware obstacle avoidance. These results underscore the decisive role of socio-spatial dependencies in strategy formation and challenge conventional paradigms that focus exclusively on individual-level behavior.

0 citationsRead paper

Measurement noise limits the advantage of nonlinear models over linear models in biomedical prediction

Jun 16, 2026

This study addresses the puzzling observation that nonlinear models often fail to outperform linear counterparts on biomedical tabular data, despite the underlying biological mechanisms being highly nonlinear. The authors demonstrate that measurement noise preferentially degrades nonlinear signal structures, thereby masking the potential advantages of complex models. To formalize this insight, they establish the first systematic link between measurement error theory and the performance gap between linear and nonlinear models, proposing a tripartite framework governed by measurement reliability, sample size, and feature representation. They further derive an excess risk identity grounded in measurement error statistics and Gaussian analysis. Empirical validation across 140 UK Biobank prediction tasks confirms that the observed performance gap strongly reflects data noise characteristics, underscoring that neither scaling data volume nor increasing model complexity can compensate for information loss induced by low measurement reliability.

0 citationsRead paper

The Umwelt Representation Hypothesis: Rethinking Universality

Apr 20, 2026

Current claims that artificial neural networks and biological brains converge on a single universal solution—termed “universality”—are limited in scope. This work proposes the Umwelt Representation Hypothesis, arguing that representational alignment arises from overlapping ecological constraints rather than convergence to a global optimum. Through empirical analyses of representational alignment across species, individuals, and artificial neural networks—combined with ecological constraint modeling and comparative neuroscience methods—the study demonstrates that representational differences are systematic and adaptive, challenging explanations based on universality. By rejecting the notion of a universal representational space, this research redefines the paradigm for model comparison, introducing alignment clusters defined within an ecological constraint space, thereby offering a novel framework for understanding the relationship between artificial and biological intelligence representations.

0 citationsRead paper

Explainable AI Methods for Neuroimaging: Systematic Failures of Common Tools, the Need for Domain-Specific Validation, and a Proposal for Safe Application

Aug 04, 2025

Widely adopted XAI methods (e.g., GradCAM, LRP) in neuroimaging lack rigorous validation and exhibit systematic localization failures, undermining model interpretability and trust. Method: We introduce the first XAI validation framework grounded in real brain MRI data: constructing prediction tasks with known ground-truth signal sources and establishing a quantifiable explanation benchmark; evaluating gradient-based methods—including SmoothGrad, GradCAM, and LRP—objectively on ~45,000 structurally annotated brain MRI scans without reliance on image perturbations. Results: SmoothGrad significantly outperforms other methods; domain mismatch—not implementation flaws—is the primary cause of failure for GradCAM and LRP. Crucially, methods making fewer simplifying assumptions (e.g., avoiding heuristic smoothing or layer-specific weighting) demonstrate superior robustness in neuroimaging. This work establishes a new empirical standard and validation paradigm for XAI in medical imaging, providing critical evidence to guide method selection and evaluation rigor in clinical AI applications.

0 citationsRead paper

Recent publications

Latest Papers

From Optimal Policies to Individual Differences: Rethinking Reinforcement Learning for Biology

Jul 17, 2026

This study addresses a key limitation in existing reinforcement learning approaches, which often overlook individual differences when modeling biological behavior, focusing instead on optimal policies or population averages. To overcome this constraint, the work introduces a biologically interpretable framework that integrates methods from multiple subfields of reinforcement learning to construct a computational model capable of generating diverse individual behaviors. By systematically synthesizing technical strategies that support behavioral diversity, the research establishes a novel paradigm for modeling individual variation in biological agents. This paradigm effectively narrows the gap between simulated and real-world biological behaviors, offering both a theoretical foundation and practical guidance for future research in biologically plausible behavior modeling.

0 citationsRead paper

Social-spatial dependencies for learning visual navigation

Jul 08, 2026

This study addresses the limitation of traditional visual navigation approaches, which often neglect the influence of social structures and spatial interactions on individual behavioral strategies within groups. The authors propose a bottom-up modeling framework that integrates deep reinforcement learning with multi-agent simulation to train agents in non-stationary dynamic environments, thereby investigating how task objectives and socio-spatial factors jointly shape the evolution of navigation strategies. Their findings demonstrate that high-quality social information can induce strategic phase transitions, successfully reproducing behavioral shifts ranging from individual navigation and group following to crowd-aware obstacle avoidance. These results underscore the decisive role of socio-spatial dependencies in strategy formation and challenge conventional paradigms that focus exclusively on individual-level behavior.

0 citationsRead paper

Measurement noise limits the advantage of nonlinear models over linear models in biomedical prediction

Jun 16, 2026

This study addresses the puzzling observation that nonlinear models often fail to outperform linear counterparts on biomedical tabular data, despite the underlying biological mechanisms being highly nonlinear. The authors demonstrate that measurement noise preferentially degrades nonlinear signal structures, thereby masking the potential advantages of complex models. To formalize this insight, they establish the first systematic link between measurement error theory and the performance gap between linear and nonlinear models, proposing a tripartite framework governed by measurement reliability, sample size, and feature representation. They further derive an excess risk identity grounded in measurement error statistics and Gaussian analysis. Empirical validation across 140 UK Biobank prediction tasks confirms that the observed performance gap strongly reflects data noise characteristics, underscoring that neither scaling data volume nor increasing model complexity can compensate for information loss induced by low measurement reliability.

0 citationsRead paper

The Umwelt Representation Hypothesis: Rethinking Universality

Apr 20, 2026

Current claims that artificial neural networks and biological brains converge on a single universal solution—termed “universality”—are limited in scope. This work proposes the Umwelt Representation Hypothesis, arguing that representational alignment arises from overlapping ecological constraints rather than convergence to a global optimum. Through empirical analyses of representational alignment across species, individuals, and artificial neural networks—combined with ecological constraint modeling and comparative neuroscience methods—the study demonstrates that representational differences are systematic and adaptive, challenging explanations based on universality. By rejecting the notion of a universal representational space, this research redefines the paradigm for model comparison, introducing alignment clusters defined within an ecological constraint space, thereby offering a novel framework for understanding the relationship between artificial and biological intelligence representations.

0 citationsRead paper

Explainable AI Methods for Neuroimaging: Systematic Failures of Common Tools, the Need for Domain-Specific Validation, and a Proposal for Safe Application

Aug 04, 2025

Widely adopted XAI methods (e.g., GradCAM, LRP) in neuroimaging lack rigorous validation and exhibit systematic localization failures, undermining model interpretability and trust. Method: We introduce the first XAI validation framework grounded in real brain MRI data: constructing prediction tasks with known ground-truth signal sources and establishing a quantifiable explanation benchmark; evaluating gradient-based methods—including SmoothGrad, GradCAM, and LRP—objectively on ~45,000 structurally annotated brain MRI scans without reliance on image perturbations. Results: SmoothGrad significantly outperforms other methods; domain mismatch—not implementation flaws—is the primary cause of failure for GradCAM and LRP. Crucially, methods making fewer simplifying assumptions (e.g., avoiding heuristic smoothing or layer-specific weighting) demonstrate superior robustness in neuroimaging. This work establishes a new empirical standard and validation paradigm for XAI in medical imaging, providing critical evidence to guide method selection and evaluation rigor in clinical AI applications.

0 citationsRead paper