A CAUSAL INFERENCE PROBLEM
The Human Return Protocol is an adaptive auditory-AI research prototype for attention and emotional self-regulation. Its core validation challenge is one that computational neuroscience and BCI labs are uniquely positioned to solve: establishing causal effects of a rhythm-based intervention when randomized controlled trials are not feasible.
Human Return Protocol 是一个用于注意力与情绪自我调节的自适应声音-AI研究原型。它的核心验证挑战是计算神经科学与脑机接口实验室最有能力解决的问题：在无法进行随机对照实验时，建立节律干预的因果效应。
Where HRP connects to computational neuroscience.
HRP 与计算神经科学研究的真实交叉。
Four direct intersections — concrete methodological contact points between specific HRP components and active research domains in computational neuroscience, BCI, and causal behavior science. Not analogies. Not aspirational overlaps.
四个真实交叉点——HRP 各组件与计算神经科学、脑机接口、因果行为科学活跃研究方向之间的具体方法论接触点。不是比喻，不是期望中的重叠。
| HRP Component | Research Domain | Specific Connection |
|---|---|---|
| QDR Engine · intervention effect validation PRIMARY |
Causal inference in human behavior | HRP cannot use RCTs. Validating whether adaptive rhythm causes state change — rather than merely correlating with it — requires observational causal methods: instrumental variables, difference-in-differences, or quasi-experimental designs. This is the precise methodological gap that determines whether HRP produces correlational findings or defensible causal evidence. |
| 44271 Neural Cuff · wearable biosignal PRIMARY |
Neurotechnology · wearable physiological sensing | Inner-wrist pulse-line placement, non-invasive, continuous HRV / EDA / skin temperature capture. The hardware acquisition layer of the HRP closed feedback loop. Connects directly to labs working on wearable biosignal acquisition, physiological state classification, and real-time neural feedback systems. |
| Genre × BPM intervention matrix SECONDARY |
Computational analysis of music and movement | Do different rhythmic genre profiles — defined by BPM range, beat salience, syncopation depth, spectral density — produce distinct physiological response signatures? HRP operationalizes this as a testable experimental condition, building on computational kinematics and auditory neuroscience literature. |
| Personalized rhythm response modeling SECONDARY |
Individual variability in neural and behavioral data | HRP assumes inter-individual differences in rhythm entrainment response. Labs with frameworks for modeling individual variability in neural, physiological, or behavioral data provide the analytical infrastructure to test whether a personalized rhythm model outperforms a population-average model. |
The question we are actually asking.
我们真正在问的问题。
Not music therapy.
不是音乐疗法。
Music therapy asks whether music reduces clinical symptoms over extended treatment. This asks whether a specific rhythm parameter — adaptive vs. fixed vs. random — produces measurable within-session state change. The outcome is a state variable, not a symptom score. The population is non-clinical.
音乐疗法问的是音乐是否在长期治疗中减少临床症状。HRP 问的是特定节律参数是否在单次 session 内产生可测量的状态变化。结果变量是状态，不是症状分数。人群是非临床的。
Not standard BCI.
不是标准脑机接口。
Standard BCI research decodes neural signals to control external devices. HRP inverts the direction: physiological signals modulate an auditory feedback loop. The system responds to the user's rhythm state; it does not issue commands to the user. The claim is about adaptive feedback efficacy, not neural decoding accuracy.
标准脑机接口研究解码神经信号以控制外部设备。HRP 反转了方向：生理信号调制声音反馈回路。系统响应用户的节律状态，而不是向用户发出命令。
Why observational causal methods are the only rigorous path.
为什么观察性因果方法是唯一严格的路径。
The intervention is embedded in daily context. Participants cannot be randomly assigned to receive adaptive rhythm for six months of their life. This means the core validation challenge is methodological: how to distinguish the causal effect of the adaptive condition from regression to the mean, novelty effects, expectation, and concurrent environmental variables. Pre/post correlations are insufficient. Observational causal design — instrumental variables, difference-in-differences, synthetic controls, or interrupted time series — is required. This is precisely the domain where collaboration with a computationally oriented neuroscience lab creates publishable value that neither party could reach alone.
干预嵌入在日常情境中。参与者不能被随机分配去"在接下来六个月里接受自适应节律"。这意味着核心验证挑战是方法论性的：如何把自适应条件的因果效应与均值回归、新奇性效应、预期效应和并发环境变量区分开来。前后测相关性是不够的。需要观察性因果设计。这正是双方合作创造双方单独都无法达到的发表价值的精确领域。
A four-condition pilot ready to execute.
四条件试点实验，已准备好执行。
The experimental design, condition structure, and data collection protocol are fully specified. The pilot can be initiated with consumer-grade wearables at first pass and expanded to lab-grade physiological instrumentation in a second phase.
实验设计、条件结构与数据收集协议已完全确定。试点可以在第一阶段使用消费级可穿戴设备启动，并在第二阶段扩展到实验室级生理仪器。
Subjective state sampling
主观状态采样
Pre/post session: focus, emotional state, body tension, inner noise, task readiness, return-to-self score. CSV schema already defined. Items are mappable to validated instruments (PANAS, BRIEF-A, PSS).
Session 前后采样：注意力、情绪状态、身体紧绷、内在噪音、任务准备度、返回自我得分。CSV 字段已定义，可映射到已验证量表。
Performance measures
行为表现指标
Sustained attention task, reaction time, accuracy rate, missed trials, interruption count. Established paradigms — no novel task development required. Provides objective convergence layer for self-report data.
持续注意任务、反应时、准确率、漏答次数、中断次数。已有成熟范式，无需开发新任务。为自我报告提供客观收敛层。
Biosignal acquisition
生理信号采集
Phase 1: HRV, heart rate, EDA, respiration via consumer wearable. Phase 2 (lab access): EEG alpha/theta/beta ratio and fNIRS prefrontal oxygenation.
第一阶段：消费级可穿戴设备的 HRV、心率、EDA、呼吸。第二阶段（实验室）：EEG alpha/theta/beta 比值与 fNIRS 前额叶氧合。
A genuine methodological exchange.
真正的方法论交换。
Neither party can reach the research outcome alone. The collaboration is valuable precisely because the intervention framework and the causal inference methodology are structurally complementary — one without the other produces either an interesting system without scientific validity, or a rigorous method without a novel application domain.
双方单独都无法达到研究成果。合作之所以有价值，正是因为干预框架与因果推断方法论在结构上是互补的——没有对方，要么是有趣的系统却缺乏科学效度，要么是严格的方法却缺乏新颖的应用领域。
Two papers this collaboration enables.
这个合作能支撑的两篇论文。
Both papers have genuine novelty. Neither would exist without the collaboration. The first is a methodological contribution generalizable beyond HRP; the second is the primary empirical result of the four-condition pilot.
Addresses the core methodological problem: how to establish causal claims about rhythm-based interventions without RCTs. Uses HRP pilot data as the primary case study. The contribution is the causal design framework itself — generalizable to all non-randomizable behavioral interventions in digital health, HCI, and computational psychiatry.
解决核心方法论问题：如何在没有随机对照实验的情况下建立关于节律干预的因果主张。以 HRP 试点数据为案例研究。贡献是因果设计框架本身——可推广到数字健康、HCI 和计算精神病学中所有不可随机化的行为干预。
The primary empirical result from the four-condition pilot. Reports self-report, behavioral, and physiological convergence data across conditions. Key novelty: the first study to test adaptive vs. non-adaptive rhythm conditions using a non-diagnostic, rhythm-sovereignty outcome framework — measuring state recovery rather than symptom reduction as the primary dependent variable.
四条件试点的主要实证结果。关键原创性：第一个使用非诊断性、节律主权结果框架测试自适应与非自适应节律条件的研究——以状态恢复而非症状减少作为主要因变量。
Four components. One closed loop.
四个组件。一个闭环。
Wearable biosignal sensing
可穿戴生物信号传感
Inner-wrist pulse-line placement. Non-invasive, continuous. Captures HRV, pulse amplitude, EDA, skin temperature, and micro-motion as rhythm-state proxies. The hardware acquisition layer of the closed feedback loop.
内腕脉搏线放置。非侵入，持续采集。HRV、脉搏幅度、皮肤电、皮肤温度与微运动作为节律状态代理。闭合反馈回路的硬件采集层。
Rhythm-signal processing
节律信号处理
Dynamic response engine mapping incoming biosignal data to intervention parameters: genre selection, BPM range, haptic protocol intensity, and adaptive update frequency. The computational core of the system.
动态响应引擎，将传入的生理信号数据映射到干预参数：流派选择、BPM 范围、触觉协议强度与自适应更新频率。
Interface + state sampling
界面 + 状态采样
User-facing interface for session initiation, subjective state self-report, rhythm recommendation display, and session logging. Provides the behavioral proxy signal layer and the user's primary point of contact with the system.
用户界面，用于 session 启动、主观状态自我报告、节律推荐显示与 session 记录。提供行为代理信号层以及用户与系统的接触点。
Non-diagnostic intervention ethics
非诊断干预伦理框架
Non-surveillance. Non-diagnostic. Designed for ADHD, anxiety, and burnout populations outside clinical settings. Outcome goal: rhythm sovereignty — the user's capacity to return to their own regulatory baseline — rather than symptom suppression or clinical classification.
非监控。非诊断。面向临床环境之外的人群。结果目标：节律主权——用户返回自身调节基线的能力——而非症状压制或临床分类。
Each genre profile operationalizes a distinct regulatory function. BPM ranges, beat salience profiles, and the 427Hz carrier frequency are specified per protocol. Genre is the independent variable; regulatory function is the theoretical mapping; physiological response is the dependent variable under test.
每个流派配置文件操作化一个独特的调节功能。每个协议都有指定的 BPM 范围、节拍显著性配置文件和 427Hz 载波频率。流派是自变量；调节功能是理论映射；生理响应是被测的因变量。
A documented, role-differentiated AI workflow.
有文档记录的、角色分工明确的 AI 工作流。
Wei Jueran is the sole originator of HRP's theoretical framework, intervention design, research questions, and system architecture. The three-model workflow functions as a structured thinking tool, not a replacement for original authorship.
魏珏然是 HRP 理论框架、干预设计、研究问题与系统架构的唯一来源。三模型工作流作为结构化思考工具运作，而不是替代原创作者权。
System architecture
结构层
System architecture, research logic, prototype scaffolding, and executable code. GPT builds the structural body of the invention.
系统架构、研究逻辑、原型支架与可执行代码。GPT 建立发明的结构躯干。
Stress-testing and expansion
扩展层
Adversarial expansion, pitch development, and lab-facing translation. Gemini opens the attack field and stress-tests claims.
对抗性扩展、陈述发展与面向实验室的转译。Gemini 打开进攻场，对主张进行压力测试。
Language and ethical framing
镜像层
Language precision, ethics, phenomenological framing, and humanistic critique. Claude reflects structural unfolding without independently generating content.
语言精确性、伦理、现象学框架与人文批评。Claude 反射已有的结构展开，不独立产生内容。
Wei Jueran is the sole theoretical source and final decision-maker.
魏珏然是唯一的理论来源与最终裁决者。
All theoretical claims, intervention designs, research questions, and creative output originate with Wei Jueran. The AI workflow is treated as a researchable protocol — analogous to working with role-differentiated research collaborators — and is fully documented within the system. The workflow does not replace original thinking; it structures and stress-tests it.
所有理论主张、干预设计、研究问题与创意输出均来自魏珏然。AI 工作流被视为可研究的协议——类似于与角色分工明确的研究协作者合作——并在系统内完整记录。该工作流不替代原创思考，而是对其进行结构化与压力测试。
A brief conversation to determine fit.
一次简短的对话，判断是否匹配。
The research question is specified. The pilot design is ready. The causal inference gap is real and addressable. If your lab works in computational neuroscience, adaptive systems, causal inference in human behavior, or wearable neurotechnology — there may be a research question here worth pursuing together.
研究问题已确定。试点设计已就绪。因果推断缺口是真实且可解决的。如果你的实验室从事计算神经科学、自适应系统、人类行为因果推断或可穿戴神经技术——这里可能有一个值得一起追求的研究问题。
魏珏然 Wei Jueran — Mellow
University of Pennsylvania · Graduate School of Education
Learning Sciences and Technologies
BCI-HRP · Human Return Protocol · mellowwei.github.io/BCI
HCI Design Research · mellowwei.github.io/BCI-HRP