January 26, 2024
Shawn Prest, Kevin Berryman
5 citations
preprint
Deep meditative deconstruction, particularly the Buddhist defabrication process and its associated phenomenology, can be understood through the active inference framework (AIF). Buddhist defabrication is a deconstructive process that drives inference ever lower in an agent's hierarchical generative model by repeatedly releasing mental tensing linked to clinging and aversion. This release corresponds to a hierarchical level-specific reduction in belief precision, allowing Buddhist concepts like equanimity and meditative stillness to be interpreted under AIF. The deconstruction process culminates in a cessation of phenomenal experience, and the states traversed may inform understanding of core-knowledge structuring and the generation of experience.
Neuroscience and biobehavioral reviews
March 1, 2026
Hagar Tal, Malcolm Wright, Shawn Prest et al.
2 citations
Computational models of advanced meditation, particularly those using Active Inference, increasingly point to precision weighting—the confidence assigned to different model parameters—as a shared mechanism that shapes shifts in experience. Early models emphasize top-down attentional modulation toward interoception or specific objects, while later models focus on layer-specific precision re-weighting within the meditator's hierarchical generative model to target more specific phenomenology. Despite progress, minimal phenomenal experiences such as nonduality and cessations remain largely unaddressed. Few models account for increased cognitive flexibility or learning from meditation, and mechanisms behind informal practice, affective processes, and compassion traditions are underexplored.
Neural computation
June 2, 2026
Shawn Prest
Meditative deconstruction—letting go of conceptual frameworks—can be modeled computationally using active inference. When an agent reduces the precision of its beliefs about hidden states at a specific hierarchical level, the phenomenology of conceptual attenuation, reduced reactivity, and shorter temporal-scale perception naturally emerges. In simulations of a facial recognition task, an agent that selects a letting-go policy when perceived affective valence becomes excessively negative can self-regulate experienced affect. The model provides a formal account of how letting go alters perception and action during meditation, offering a computational perspective on equanimity, stillness, and affect regulation.