Cai & Cai's 'Canxianization' theory explains why unfinished thoughts keep returning

Hengjin Cai and Tianqi Cai's new paper on arXiv (arXiv:2605.12543) introduces 'Canxianization,' a formal model explaining why some unfinished tasks or unresolved problems repeatedly intrude on conscious thought long after their triggering conditions are gone. The researchers distinguish this phenomenon from the well-known Zeigarnik effect, emotional arousal, memory strength, curiosity, prediction error, and intrusive thought. They define a perturbation as 'canxianized' when it is attributed to the self-world boundary, value-marked, blocked from causal or action closure, and metacognitively coupled to the self-model.

The paper introduces two quantitative metrics: a Recurrent Priority Index (measuring how often a thought returns) and a Canxian Update Index (tracking whether the recurrence is productive or pathological). A critical innovation is 'Cold Canxianization' — recurrence driven purely by structural incompleteness rather than affective arousal. The authors propose Reset Resistance and Stake Transfer tests that could be applied to artificial systems (like AI agents) to determine whether they exhibit similar closure-seeking behavior. The work frames Canxianization not as memory persistence but as 'failed self-world repair,' suggesting the brain prioritizes unresolved self-relevant patterns to protect the integrity of the self-model. This could have implications for understanding rumination, PTSD, persistent curiosity, and even design of goal-oriented AI systems.