Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
The paper identifies reward misspecification as a fundamental flaw in closed-loop EEG and biofeedback systems, where optimizing for measurable proxy signals like calm EEG or HRV coherence leads to unintended strategies unrelated to the intended benefit. They formalize this issue and categorize failure modes as proxy mismatch, strategy shortcutting, and transfer failure, illustrating these with examples from existing devices. To address this, they propose a design framework centered on Tier-1 measurable targets, negative-only cueing, temporal separation of EEG and somatic features, and transfer to unassisted practice as the key success metric.
Meditation wearables often fail because they incentivize users to game the system, optimizing for easily measured signals instead of actual mental well-being.
Consumer EEG headbands, HRV biofeedback devices, and closed-loop neurostimulation systems share a fundamental design flaw: they reward measurable proxy signals rather than the outcomes they claim to produce. When a user optimises for calm EEG, HRV coherence, or breathing resonance, their brain learns to produce those signals through whatever strategy is most efficient, including strategies unrelated to the intended benefit. We formalise this as reward misspecification: the policy maximising proxy reward R_proxy is not the policy maximising true intended outcome V_target. This produces three failure modes: proxy mismatch, strategy shortcutting, and transfer failure. We review how existing devices including Muse, HeartMath, Unyte IOM2, and clinical neurofeedback systems instantiate these failures. We introduce a four-tier measurability taxonomy distinguishing reliably measurable wearable targets (Tier 1) from targets that are currently or possibly structurally unmeasurable (Tiers 3 and 4), and show that most devices make implicit Tier 3 and 4 claims. We propose a design framework that avoids all three failure modes: single Tier-1 target (mind-wandering onset via EEG), negative-only cueing, temporal separation of fast EEG and slow somatic feature streams, and transfer to unassisted practice as the only success criterion. No current product meets all four criteria. The framework has direct implications for the design, evaluation, and regulation of cognitive and contemplative wearables.