The modern connectivist framing of neural computation emphasizes the primacy of synaptic communication at the risk of neglecting the influence of the surrounding neuromodulatory environment --- a neuron's 'biophysical context.' Decades of experimental work has established two views of neuromodulatory (NMs) influence: 1) NMs significantly alter circuit dynamics and 2) NMs gate synaptic plasticity, acting as a 'third factor' in learning. Here, we unify these perspectives, proposing that neuromodulation via volume transmission implements a powerful computational principle: context factorization. We derive an endogenously neuromodulated Recurrent Neural Network (e-nmRNN) from a rate reduction of NM release, showing how NM concentrations dynamically factorize network connectivity. This framework reveals how multiplicative NM gating distinctly influences dynamical regimes compared to additive input. Crucially, this context factorization enables targeted online credit assignment: learning rules derived for the e-nmRNN are naturally gated by NM concentrations, localizing updates to relevant contexts. We demonstrate that e-nmRNN dynamics can learn to approximate gradient descent, facilitating rapid in-context adaptation akin to meta-learning. Empirically, e-nmRNNs achieve strong compositional generalization in sequence-to-sequence tasks, outperforming baselines and exhibiting greater hyperparameter robustness. Furthermore, when trained on complex multitasking benchmarks, e-nmRNNs develop emergent properties mirroring biological observations, including modularity, cell-type specialization based on NM release, and distinct neuromodulatory timescales encoding task context. The model's interpretability allows us to reverse engineer these emergent structures. Notably, in reinforcement learning tasks, the e-nmRNN learns to encode context and signals like Reward Prediction Error (RPE) within its neuromodulator dynamics, demonstrating a mechanism for RPE-gated online credit assignment essential for learning how to learn. By bridging biophysical mechanisms with computational principles and empirical validation, our work presents e-nmRNNs as a performant, interpretable model for understanding the computational role of neuromodulation in flexible and compositional learning.

Learning a compact representation of history is critical for planning and generalization in partially observable environments. While meta-reinforcement learning (RL) agents can attain near Bayes-optimal policies, they often fail to learn the compact, interpretable Bayes-optimal belief states. This representational inefficiency potentially limits the agent's adaptability and generalization capacity. Inspired by predictive coding in neuroscience---which suggests that the brain predicts sensory inputs as a neural implementation of Bayesian inference---and by auxiliary predictive objectives in deep RL, we investigate whether integrating self-supervised predictive coding modules into meta-RL can facilitate learning of Bayes-optimal representations. Through state machine simulation, we show that meta-RL with predictive modules consistently generates more interpretable representations that better approximate Bayes-optimal belief states compared to conventional meta-RL across a wide variety of tasks, even when both achieve optimal policies. In challenging tasks requiring active information seeking, only meta-RL with predictive modules successfully learns optimal representations and policies, whereas conventional meta-RL struggles with inadequate representation learning. Finally, we demonstrate that better representation learning leads to improved generalization. Our results strongly suggest the role of predictive learning as a guiding principle for effective representation learning in agents navigating partial observability.

In many scientific disciplines, we are interested in inferring the nonlinear dynamical system underlying a set of observed time series, a challenging task in the face of chaotic behavior and noise. Previous deep learning approaches toward this goal often suffered from a lack of interpretability and tractability. In particular, the high-dimensional latent spaces often required for a faithful embedding, even when the underlying dynamics lives on a lower-dimensional manifold, can hamper theoretical analysis. Motivated by the emerging principles of dendritic computation, we augment a dynamically interpretable and mathematically tractable piecewise-linear (PL) recurrent neural network (RNN) by a linear spline basis expansion. We show that this approach retains all the theoretically appealing properties of the simple PLRNN, yet boosts its capacity for approximating arbitrary nonlinear dynamical systems in comparatively low dimensions. We employ two frameworks for training the system, one combining BPTT with teacher forcing, and another based on fast and scalable variational inference. We show that the dendritically expanded PLRNN achieves better reconstructions with fewer parameters and dimensions on various dynamical systems benchmarks and compares favorably to other methods, while retaining a tractable and interpretable structure.

Virtual reality (VR) sickness is one of the obstacles hindering the growth of the VR market. Different VR contents may cause various degree of sickness. If the degree of the sickness can be estimated objectively, it adds a great value and help in designing the VR contents. To address this problem, a novel content-based VR sickness assessment method which considers both the perception prior and hybrid temporal features is proposed. Based on the perception prior which assumes the user's field of view becomes narrower while watching videos, a Gaussian weighted optical flow is calculated with a specified aspect ratio. In order to capture the dynamic characteristics, hybrid temporal features including horizontal motion, vertical motion and the proposed motion anisotropy are adopted. In addition, a new dataset is compiled with one hundred VR sickness test samples and each of which comes along with the Discomfort Scores (DS) answered by the user and a Simulator Sickness Questionnaire (SSQ) collected at the end of test. A random forest regressor is then trained on this dataset by feeding the hybrid temporal features of both the present and the previous minute. Extensive experiments are conducted on the VRSA dataset and the results demonstrate that the proposed method is comparable to the state-of-the-art method in terms of effectiveness and efficiency.