Editorial: Advancements in neural coding: sensory perception and multiplexed encoding strategies

Understanding how the nervous system encodes and interprets sensory information remains a central challenge in neuroscience. Classical models have often conceptualized sensory pathways as hierarchical and largely feedforward systems that relay stimulus features with increasing abstraction (Felleman Riesenhuber Kamaleddin et al., 2021;Kayser et al., 2009;Panzeri et al., 2010). The contributions gathered in this Research Topic collectively advance this perspective by integrating theoretical, computational, and applied approaches to sensory coding across biological and artificial systems.A key conceptual contribution is the notion that sensory signals are inherently context-dependent representations rather than fixed encodings. The perspective by Ethier and colleagues introduces the concept of sensory polysemia, proposing that identical neural activity patterns can convey different meanings depending on behavioral, emotional, hormonal, and motivational states.Drawing on the trigeminal system, the authors show how inhibitory circuits gate sensory transmission at early stages. This enables flexible interpretation of tactile inputs. This framework challenges strictly feedforward accounts of perception and instead emphasizes that meaning is constructed through interactions between incoming stimuli and internal state. Such a view aligns with broader theories of predictive and active sensing, in which perception is shaped by expectations and task demands (Clark, 2013;Friston, 2010).Complementing this conceptual advance, several contributions explore mechanistic and computational implementations of multiplexed encoding. Yedutenko et al. present a detailed investigation of time-difference encoders (TDEs) in spiking neural networks, demonstrating how motion information can be encoded through temporal correlations in event-based sensory streams.Their introduction of a three-point architecture (TDE-3), incorporating inhibitory input, addresses a key limitation of earlier models by restoring direction selectivity in complex environments. This work further highlights that inhibitory interactions are not merely suppressive but play an active computational role in disambiguating sensory signals, echoing biological circuit principles (Isaacson Kamaleddin et al., 2022). Additionally, bridging scales from single-neuron dynamics to large-scale network function continues to be a critical frontier. Finally, translating insights from controlled experimental paradigms to real-world applications, including robotics and clinical neurotechnology, will require integrating robustness, adaptability, and interpretability.Moving beyond static and unidimensional representations, the field increasingly recognizes that sensory and neural signals are dynamically interpreted, contextually modulated, and multiplexed across dimensions. By combining theoretical frameworks, computational models, and applied methodologies, the contributions presented here provide a more unified account of how neural systems encode information and how these principles can be leveraged in next-generation intelligent systems.