逐项查阅:中文端点与补表原说明
共50项。展开后可查原始英文端点、作者操作说明、量尺判读与补表行号。
Standards · 个人标准是否相符CPM / 评价 · 双极
违背个人标准 → 符合个人标准
补表原始端点:NotMyStandards → MyStandards
补表短说明为不相符,但操作说明明确上端符合、下端违背;据操作说明判为双极。
A situation is congruent with your standards or ideas, if what is happending and how people are acting is in line with your personal values and you would consider it as fair. Move the gauge up the more the situation alignes with your standards and ideas and down if what is happening violates them. You may use the middle of the scale for situations that are neutral with regards to your standards and ideas.
官方补表第 3 行;扫描组片后验证:包含。PleasantSelf · 对自己的愉悦性CPM / 评价 · 双极
对我不愉快 → 对我愉快
补表原始端点:Unpleasant Me → Pleasant For Me
按补表端点及以下操作说明判读。
Pleasant discribes a situation that was generally positive for you. Move the gauge up the more pleasant you perceive the situation and down as the situation is more unpleasant. You may use the middle of the scale for situation that are neither positive or negative.
官方补表第 4 行;扫描组片后验证:包含。SocialNorms · 是否符合社会规范CPM / 评价 · 双极
违背社会规范 → 符合社会规范
补表原始端点:NotSocialNorms → SocialNorms
下端虽写NotSocialNorms,操作说明明确是违反规范;中间为中性。
A situation is congruent with social norms, if what is happening and how people are acting is in line with your cultures expectations regarding social rules and behaviour. Move the gauge up the more the situation is conform with your laws or social norms and down if what is happening violates these ideas. You may use the middle of the scale for situations that are neutral with regards to laws and social norms.
官方补表第 5 行;扫描组片后验证:包含。PleasantOther · 对他人的愉悦性CPM / 评价 · 双极
对他人不愉快 → 对他人愉快
补表原始端点:UnpleasantOthers → PleasantOthers
按补表端点及以下操作说明判读。
Pleasant for somebody else discribes a situation that you perceive as generally positive for somebody else. Move the gauge up the more pleasant you perceive the situation is for somebody else and down as the situation is more unpleasant for them. You may use the middle of the scale for situation that are neither positive or negative. If you find conflict with different characters in the movie focus on the character you feel most invested in.
官方补表第 6 行;扫描组片后验证:包含。GoalsOther · 是否满足他人目标CPM / 评价 · 单极
不满足其目标 → 满足其目标
补表原始端点:NotOthersGoals → MeetOthersGoals
按补表端点及以下操作说明判读。
Important and relevant for the goals or needs of somebody is a situation that will help somebody achieve their wants and needs. Move the gauge up in situations that are important and relevant for somebodys goals and down for situations that are irrelevant. If you find conflict with different characters in the movie focus on the character you feel most invested in.
官方补表第 7 行;扫描组片后验证:包含。Controlled · 事件可控程度CPM / 评价 · 单极
不可控/较少可控 → 可控/更多可控
补表原始端点:Uncontrolled → EventControlled
按“可控性的多少”判为单极;不可控/可控也可被描述为相对状态,作者未给正式单/双极标签。
Controlled is an event that could have been averted or stopped. Move the gauge up the more controlled you perceive the situation and down as the situation is less controlled i.e. uncontrolled.
官方补表第 8 行;扫描组片后验证:包含。Predictable · 事件可预测程度CPM / 评价 · 单极
不可预测 → 可预测
补表原始端点:Unpredictable → EventPredictable
按“可预测性的多少”判为单极;这是连续属性,非两种独立情绪。
An event is predictable and could have been forseen. Move the gauge up the more predictable you perceive the situation to be an down for situations that are unpredictable.
官方补表第 9 行;扫描组片后验证:包含。Suddenly · 事件突然程度CPM / 评价 · 单极
不突然 → 突然
补表原始端点:Not Suddenly → EventSuddenly
按补表端点及以下操作说明判读。
A sudden event occurred without much lead up or warning, the situation changed from one moment to another. Move the gauge up the more surprising and sudden an event occurred.
官方补表第 10 行;扫描组片后验证:包含。Agent · 事件的原因来源CPM / 评价 · 双极
偶然导致 → 人类导致
补表原始端点:ChanceCaused → HumanCaused
偶然原因与人类原因的来源对照,不是正负情绪;说明允许混合或既非两者时使用中间。
Rate the degree to which the event is caused by a human agent (including personified animals and machines) versus caused by chance. Move the gauge up the more you think the event was caused by a human agent and down the more it is caused by change. Use the middle of the scale if influences are even or if the cause was neither human nor chance.
官方补表第 11 行;扫描组片后验证:包含。Urgency · 情境的紧迫程度CPM / 评价 · 单极
不紧迫 → 紧迫
补表原始端点:Not Urgent → Situation Urgent
按补表端点及以下操作说明判读。
There is urgency in the situation to act. Move up the gauge to indicate that there is urgency in the situation, move it down if the situation is not urgent.
官方补表第 12 行;扫描组片后验证:包含。Lips · 自己抿紧嘴唇CPM / 表达 · 单极
嘴唇放松 → 抿紧嘴唇
补表原始端点:Relax Lips → Press Lips
按补表端点及以下操作说明判读。
You press your lips together. Move the gauge up to indicate pressing your lips together. When your lips are relaxed move the gauge down.
官方补表第 14 行;扫描组片后验证:包含。Tears · 自己眼中含泪CPM / 表达 · 单极
眼中无泪 → 眼中含泪
补表原始端点:Dry Eyes → Tears In My Eyes
按补表端点及以下操作说明判读。
You have tears in your eyes, feel your eyes tearing up. Move the gauge up as you perceive tearing in your eyes. Move it down if your eyes are dry, that is no more tears than usual.
官方补表第 15 行;扫描组片后验证:包含。Eyebrows · 自己抬眉CPM / 表达 · 单极
眉毛放松 → 抬起眉毛
补表原始端点:BrowsRelaxed → BrowsRaised
按补表端点及以下操作说明判读。
Move the gauge up as you feel your eyebrows go up. Move it down as your eyebrows are relaxed in their normal position.
官方补表第 17 行;扫描组片后验证:包含。Smile · 自己微笑CPM / 表达 · 单极
没有微笑 → 微笑
补表原始端点:Did Not Smile → Smiled
按补表端点及以下操作说明判读。
Move the gauge up as you observe yourself smiling. Move it down as you are not smiling.
官方补表第 18 行;扫描组片后验证:包含。Frown · 自己皱眉CPM / 表达 · 单极
没有皱眉 → 皱眉
补表原始端点:Did Not Frown → Frowned
按补表端点及以下操作说明判读。
Move the gauge up as you observe yourself frowning. Move it down as you are not frowning.
官方补表第 19 行;扫描组片后验证:包含。Stop · 想制止事件CPM / 动机 · 单极
不想制止 → 想制止
补表原始端点:Don't Want Stop → Want Stop Event
按补表端点及以下操作说明判读。
Move the gauge up as you feel the urge to intervene and stop what is happening. Move it down if you do not feel an urge to stop what is happening.
官方补表第 22 行;扫描组片后验证:包含。Undo · 想撤销已经发生的事CPM / 动机 · 单极
不想撤销 → 想撤销
补表原始端点:Don't Want Undo → Want Undo Event
按补表端点及以下操作说明判读。
Move the gauge up as you feel the desire to undo what is happening. Move it down as you do not feel an urge to undo what is happening.
官方补表第 23 行;扫描组片后验证:包含。Repeat · 想延续或重现事件CPM / 动机 · 单极*
不想延续/重现 → 想延续/重现
补表原始端点:NotWantRepeat → WantLastRepeat
暂按端点判为单极。说明同时把中间称为无所谓,故不能仅据两端确定心理学上严格的单极性。
Move the gauge up as you want the ongoing situation to last or be repeated. Move it down if you don't want it to last or be repeated. You may use the middle of the scale if you are indifferent with respects to the situation to last or be repeated.
官方补表第 24 行;扫描组片后验证:包含。Oppose · 想反对他人或事物CPM / 动机 · 单极
不想反对 → 想反对
补表原始端点:NotWantOppose → Want To Oppose
按补表端点及以下操作说明判读。
Move the gauge up if you want to oppose someone, something or the situation as a whole. Move the gauge down if you do not feel an urge to oppose someone or something.
官方补表第 25 行;扫描组片后验证:包含。Attention · 注意被吸引的程度CPM / 动机 · 双极*
不感兴趣 → 注意被吸引
补表原始端点:Disinterested → Captured
暂按“不感兴趣↔注意被吸引”判为状态对照双极;说明末句误提重复事件,疑为复制残留。
Move the gauge up as you feel more motivated to pay attention to what was going on. Move it down if you feel disinterested in paying attention. You may use the middle of the scale if you feel indifferent with respect to the situation being repeated.
官方补表第 26 行;扫描组片后验证:包含。Tackle · 想做点什么应对事件CPM / 动机 · 单极
不想采取行动 → 想采取行动
补表原始端点:NotDoSomething → WantDoSth
按补表端点及以下操作说明判读。
Move the gauge up if you feel like tackling the situation and doing something. Move it down as you do not feel an urge to takle the situation.
官方补表第 27 行;扫描组片后验证:包含。Command · 想指挥或控制他人CPM / 动机 · 单极
不想指挥 → 想指挥
补表原始端点:Don'tWant → WantCommand
按补表端点及以下操作说明判读。
Move the gauge up as you want command others. Move it down as you do not feel an urge to be in command of others.
官方补表第 28 行;扫描组片后验证:包含。Support · 想帮助或陪伴他人CPM / 动机 · 单极*
不想帮助 → 想帮助
补表原始端点:Don'tWantHelp → WantHelp
暂按“无帮助愿望→有帮助愿望”判为单极;说明亦把中间称为中性,存在解释张力。
Move the gauge up if you want someone to be there to provide help or support. Move it down as you do not feel the desire for someone to be there for support or help. You may use the middle of the scale if you are indifferent with regards to someone being there to provide help and support.
官方补表第 29 行;扫描组片后验证:包含。Move · 想活动身体CPM / 动机 · 单极
不想动 → 想动
补表原始端点:Don't Want Move → Want To Move
按补表端点及以下操作说明判读。
Move the gauge up as you feel like you want to move. Move it down as you do not feel an urge to move.
官方补表第 30 行;扫描组片后验证:包含。Care · 想照顾他人CPM / 动机 · 单极
不想照顾 → 想照顾
补表原始端点:NotCareFor → WantToCareFor
按补表端点及以下操作说明判读。
Move the gauge up if you want to take care of another person or cause. Move it down as you do not feel the urge to take care of another person or cause.
官方补表第 31 行;扫描组片后验证:包含。Bad · 自己感觉难受CPM / 主观感受 · 单极*
较少难受† → 更难受†
补表原始端点:Feel Bad → Feel Bad
补表Lower Pole和Upper Pole均为Feel Bad,原文自相矛盾。本图按详细说明“向上更难受、向下较少难受”显示中文端点,未把推断冒充原始端点。
Move the gauge up as you feel bad. Move it down as you feel less bad.
官方补表第 32 行;扫描组片后验证:包含。Good · 自己感觉良好CPM / 主观感受 · 单极
不感觉良好 → 感觉良好
补表原始端点:Not Feel Good → Feel Good
按补表端点及以下操作说明判读。
Move the gauge up as you feel good. Move it down as you feel less good.
官方补表第 33 行;扫描组片后验证:包含。Calm · 自己平静或躁动CPM / 主观感受 · 双极
躁动不安 → 平静
补表原始端点:Feel Restless → Feel Calm
按补表端点及以下操作说明判读。
Calm is the response to quiet and peaceful conditions. For example, you may experience calm when you are in a safe situation. Move the gauge up as you feel more calm. Move it down if you feel more restless.
官方补表第 34 行;扫描组片后验证:包含。Strong · 自己有力量或无力CPM / 主观感受 · 双极
无力 → 有力量
补表原始端点:Feel Weak → Feel Strong
按补表端点及以下操作说明判读。
Move the gauge up as you feel strong. Move it down as you feel weak. You may use the middle of the scale to indicate feeling neither particularly strong nor weak.
官方补表第 35 行;扫描组片后验证:包含。IntenseEmotion · 情绪体验的强烈程度CPM / 主观感受 · 单极
没有强烈情绪 → 情绪强烈
补表原始端点:No Intense Emoti → Intense Emotion
按补表端点及以下操作说明判读。
Move up the gauge as you experience an intense emotional state. Move it down as you do not feel intense emotions.
官方补表第 36 行;扫描组片后验证:包含。Alert · 自己清醒或疲惫CPM / 主观感受 · 双极
疲惫 → 清醒/警觉
补表原始端点:Feel Tired → Feel Alert
按补表端点及以下操作说明判读。
Move the gauge up as you feel alert and attentive. Move it down as you feel tired. You may use the middle of the scale to describe a level of alertnes that is 'normal' for you.
官方补表第 37 行;扫描组片后验证:包含。AtEase · 自己放松或紧张CPM / 主观感受 · 双极
紧张 → 放松自在
补表原始端点:Feel Nervous → Feel At Ease
按补表端点及以下操作说明判读。
Move the gauge up as you feel at ease. Move it down as you feel more nervous. You may use the middle of the scale to indicate a state of neither feeling particularly at ease nor feeling nervous.
官方补表第 38 行;扫描组片后验证:包含。Muscle · 肌肉的紧张程度CPM / 生理感受 · 单极*
肌肉放松 → 肌肉紧张
补表原始端点:Muscles Relax → Muscles Tense
按“放松基线→紧张增加”暂判单极;详细说明又允许“中间为正常”,严格单/双极性质需额外确认。
Move the gauge up as you experience muscle tensions in your whole body. Move it down as you experience your muscles relaxing. You may use the middle of the scale for what you perceive as your normal level of muscle tension.
官方补表第 39 行;扫描组片后验证:包含。Heartrate · 感觉心跳变慢或变快CPM / 生理感受 · 双极
心跳更慢 → 心跳更快
补表原始端点:Heartbeat Slower → Heartbeat Faster
这是对自己心跳的主观感觉,不是仪器测得的脉搏。该项未纳入扫描组48项片后验证。
Move the gauge up as you experience your heartbeat going faster. Move it down as you experience your heartbeat going slower. You may use the middle of the scale for what you perceive as your normal heartrate.
官方补表第 40 行;扫描组片后验证:未包含。Throat · 感觉喉咙哽咽CPM / 生理感受 · 单极
喉咙正常通畅 → 喉咙哽咽
补表原始端点:Throat Clear → Lump In Throat
按补表端点及以下操作说明判读。
Move the gauge up as you feel a lump in your throat. Move it down as your throat feels clear as usual.
官方补表第 42 行;扫描组片后验证:包含。Stomach · 感觉胃部不适CPM / 生理感受 · 单极
胃部正常 → 胃部不适
补表原始端点:Stomach Normal → Stomach Unwell
按补表端点及以下操作说明判读。
Move the gauge up as you feel discomfort in your stomach. Move it down as your stomach feels comfortable as normal.
官方补表第 43 行;扫描组片后验证:包含。Warm · 自己感觉冷或暖CPM / 生理感受 · 双极
感觉冷 → 感觉暖
补表原始端点:Felt Cold → Felt Warm
这是主观冷热感觉。该项未纳入扫描组48项片后验证。
Move the gauge up as you felt physically warm. Move it down if you feel physically cold. You may use the middle of the scale to describe feeling at a temperature that is 'normal' for you.
官方补表第 44 行;扫描组片后验证:未包含。Anger · 愤怒离散情绪 · 单极
不愤怒 → 愤怒
补表原始端点:Not Angry → Angry
按补表端点及以下操作说明判读。
Anger is a feeling of irritation or rage towards another person or situation. Move up the gauge as you feel increasing anger. Move it down as you do not feel angry.
官方补表第 45 行;扫描组片后验证:包含。Guilt · 内疚离散情绪 · 单极
不内疚 → 内疚
补表原始端点:Not Guilty → Guilty
按补表端点及以下操作说明判读。
Guilt is the response when a person regrets having violated agreements, principles, or values. For example, you may experience guilt when some behaviours or situations violate your own values or moral code. Move the gauge up as you feel guity. Move it down as you feel less guilty.
官方补表第 46 行;扫描组片后验证:包含。WarmHeartedness · 温情/亲切感离散情绪 · 单极
没有温情 → 有温情
补表原始端点:NotWarmHeart → WarmHearted
按补表端点及以下操作说明判读。
Warm-hearted is a feeling of sympathy, affection, kindness, and cordiality for situations or people. For example, you may experience warm-hearted when you witness an event that you find very positive and comforting. Move the gauge up as you feel more warm-hearted. Move it down if you do not feel warm-hearted towards the situation or the people in it.
官方补表第 47 行;扫描组片后验证:包含。Disgust · 厌恶离散情绪 · 单极
不厌恶 → 厌恶
补表原始端点:Not Disgusted → Disgusted
按补表端点及以下操作说明判读。
Disgust is the response to sights, situations, and actions that we do not accept. For example, you may experience disgust when someone behaves in a way that you disapprove. Move the gauge up as you feel disgusted. Move it down as you do not feel disgusted.
官方补表第 48 行;扫描组片后验证:包含。Happiness · 快乐离散情绪 · 单极
不快乐 → 快乐
补表原始端点:Not Happy → Happy
定义句提到sad,但显示端点是Not Happy→Happy;不据此改成悲伤↔快乐。
Happiness is a feeling of joy opposite to feeling sad. Move the gauge up as you feel more happy. Move it down as you feel less happy.
官方补表第 49 行;扫描组片后验证:包含。Fear · 恐惧离散情绪 · 单极
不恐惧 → 恐惧
补表原始端点:Not Fearful → Fearful
按补表端点及以下操作说明判读。
Fear is the response to concrete physical and psychological threats. For example, you may experience fear when your/someone’s life, health, or well-being is threatened. Move up the gauge as you feel fearful. Move it down as you feel less fearful.
官方补表第 50 行;扫描组片后验证:包含。Regard · 敬重(含轻蔑一端)离散情绪 · 双极
轻蔑 → 敬重
补表原始端点:Contempt → Regard
按补表端点及以下操作说明判读。
Regard is a feeling of respect and sympathy towards another person. Opposite to regard; contempt is a feeling of moral superiority, disrespect, and aversion towards another person. Move the gauge up as you feel regard. Move it down as you feel contempt. You may use the middle of the scale if you feel neither regard nor contemt.
官方补表第 51 行;扫描组片后验证:包含。Anxiety · 焦虑离散情绪 · 单极
不焦虑 → 焦虑
补表原始端点:Not Anxious → Anxious
按补表端点及以下操作说明判读。
Anxiety is a feeling of tension, apprehension, worry, and nervousness. For example, you may experience anxiety when you worry about something that is happening or might happen in the future. Move up the gauge as you feel more anxious. Move it down as you feel less anxious.
官方补表第 52 行;扫描组片后验证:包含。Satisfaction · 满意(含不满一端)离散情绪 · 双极
不满意 → 满意
补表原始端点:Dissatisfied → Satisfied
按补表端点及以下操作说明判读。
Satisfaction is the response to the fulfilment of our wishes, expectations, and needs. Move the gauge up as you feel more satisfied. Move it down as you feel dissatisfied. You may use the middle of the scale to describe a neutral state of being neither satisfied nor dissatisfied.
官方补表第 53 行;扫描组片后验证:包含。Pride · 自豪(含羞耻一端)离散情绪 · 双极
羞耻 → 自豪
补表原始端点:Ashamed → Proud
按补表端点及以下操作说明判读。
Pride is a feeling of pleasure or satisfaction with ones achievements. Shame is a feeling of humiliation or distress when a person is conscious of their wrongdoings or faults. Move the gauge up as you feel more proud. Move it down as you feel more ashamed. You may use the middle of the scale if you feel neither pride nor shame.
官方补表第 54 行;扫描组片后验证:包含。Surprise · 惊讶离散情绪 · 单极
不惊讶 → 惊讶
补表原始端点:Not Surprised → Surprised
按补表端点及以下操作说明判读。
Surprise is the response to sudden expected events. For example, you may experience surprise when you are in unusual situations or someone acts unexpectedly. Move the gauge up as you feel more surprised. Move it down as you feel less surprise.
官方补表第 55 行;扫描组片后验证:包含。Love · 爱离散情绪 · 单极
没有爱 → 有爱
补表原始端点:No Love → Love
按补表端点及以下操作说明判读。
Love is a strong feeling of liking some people very much and being attached to them. For example, you may experience love when you like a person, a child, or a lover very much. Move the gauge up as you feel more love. Move it down as you feel less love.
官方补表第 56 行;扫描组片后验证:包含。Sad · 悲伤离散情绪 · 单极
不悲伤 → 悲伤
补表原始端点:Not Sad → Sad
按补表端点及以下操作说明判读。
Sadness is the response to the loss of an object or a person to which you are attached. Move the gauge up as you feel more sad. Move it down as you do not feel sad
官方补表第 57 行;扫描组片后验证:包含。后续研究:先区分预测对象
相同数据集可以回答完全不同的问题。“预测情绪评分”“预测一个人的焦虑问卷分数”“预测电影画面”不能都叫作情绪解码。
| 研究路线 | 典型输入 → 输出 | 本轮核对的工作 |
|---|
| 情绪的组织与表征 | CPM评分 → 情绪评分;脑活动 → 情绪类别/关系 | 成分分析论文、情绪认知地图论文 |
| 动态与条件差异 | 情绪选段 → 连接梯度 → 个体差异;情绪变化 ↔ 脑状态变化 | 功能梯度、脑—心耦合;另列2份会议摘要 |
| 测量泛化与视觉建模 | 换电影检验脑同步;电影帧 ↔ 视觉区信号 | 跨电影ISC、视觉编码—解码 |
查找范围:截至2026-10-21,以正式论文与2024数据预印本的OpenAlex被引条目为线索,再按Emo-FilM名称/数据集编号检索并回到原文。下面是已核到具体复用方法的6项工作和2份会议摘要,另列一篇实际用评分作图的综述。不是声称穷尽全部引用;正式版全文未取得时,明确标注公开预印本版本。
01 / 情绪的组织与表征
Dissecting appraisal and multicomponential features of emotion
2026;方法细节据2024公开预印本 · 正式论文已发表;全文核查到预印本
情绪名称背后,是否有可区分的评价、动机和身体感受组合?
用了哪些数据
14部电影的50项群体共识评分:37项CPM作输入,13项离散情绪作预测对象;约9,600个秒级观测。不使用fMRI或仪器生理记录。
→
如何分析
先比较13项的时间相关与层次聚类;再按情绪评分加权汇总37项CPM,形成每种情绪的成分轮廓。预设3因子检验效价/唤醒/力量结构。最后用分类模型分别预测单项评分高低、13类中最高评分项,以及四个较粗情绪簇,并做留一电影验证。
→
得到什么结果
评分间可见效价与自我/他人相关的组织;CPM项目可跨电影预测情绪评分,评价项目普遍有信息,动机等补充区分。预印本中,13类任务的留一电影准确率43.93%,多数类基线25.64%;混合电影的10折结果94.42%,不可混报。
解读边界:这些是群体评分之间的统计联系,不能证明评价在时间上引起了情绪。3因子是事先设定,不能说数据发现“情绪恰好只有3维”。数字据预印本,未冒充正式版复核值。
核对位置:预印本 Methods §2.2;Results §3.1–3.5;Discussion pp.17–22。正式版摘要与EPFL书目已核对。
原始论文/作者摘要 ↗ · 本轮另核对的公开来源 ↗
02 / 情绪的组织与表征
Map-like representations of emotion knowledge in hippocampal-prefrontal systems
2026 · Nature Communications · 正式全文:方法、结果与讨论已核对
大脑只是区分情绪类别,还是也组织了类别之间的关系,像一张认知地图?
用了哪些数据
29名扫描被试(因缺失影片排除1名)的fMRI;13项离散情绪,加Good、Bad、Calm、AtEase这4项主观感受。不使用全部50项。
→
如何分析
用偏最小二乘回归从海马、内嗅皮层、vmPFC等区域预测类别评分、二维情绪空间轨迹等,留一电影验证。另以评分构造的环境训练Tolman–Eichenbaum Machine(TEM),比较其概念/位置联合编码p与关系编码g能否从脑信号读出。
→
得到什么结果
海马相对更支持具体类别与细粒度联合编码;vmPFC更支持情绪空间轨迹及较大尺度关系编码。海马类别评分的平均Fisher z≈0.0525,二维位置≈0.0428;这是预测相关,绝不是“识别准确率”。
解读边界:TEM是计算模型,不是直接记录到的神经元。地图式表征是模型比较支持的解释;不能说已经证明脑内存在本研究绘出的二维地图,作者也承认二维简化了情绪。
核对位置:Methods: Emotion ratings / Multivariate decoding;Results Fig.2–5;Discussion limitations。
原始论文/作者摘要 ↗
03 / 情绪与脑活动的动态/条件差异
Individual Differences of Cortical and Subcortical Emotion-Informed Functional Gradients
2026 · Human Brain Mapping · 正式全文:方法、结果与讨论已核对
用什么情绪场景的脑数据,最容易看出焦虑或开放性等个体差异?
用了哪些数据
30人的电影与静息态fMRI、50项共识评分,以及15项个体差异问卷分数。50项先按既有因子载荷归并为效价、唤醒、力量的时间序列。
→
如何分析
按情绪分数的四分位选择电影时间段;由414脑区计算功能连接,扩散嵌入为功能梯度,提取各脑区到梯度中心的距离。用Lasso及嵌套被试交叉验证预测问卷,并比较静息、整段电影和按情绪选段。
→
得到什么结果
电影梯度较适合预测状态焦虑,静息梯度较适合预测开放性;按情绪选段后两者均可改善。焦虑的最佳平均预测相关约0.33(整段电影0.26),开放性约0.32(整段电影0.03);不同目标适合不同场景。
解读边界:预测目标是人的问卷分数,不是当前时刻的50项情绪准确率。样本30人,最佳选段结果属探索性;不能总结为“电影永远优于静息”,也不是因果操纵情绪。
核对位置:Methods §2.1.2、§2.2.4–2.2.6;Results §3.3、§3.7–3.8。
原始论文/作者摘要 ↗
04 / 情绪与脑活动的动态/条件差异
Brain-heart coupling shapes large scale brain dynamics
2025 · bioRxiv预印本 · 作者公开全文:外部验证方法与结果已核对
心率变化与脑网络整合/分离的关系,在电生理和BOLD中是否相同?
用了哪些数据
主研究另有颅内EEG与另一组fMRI。Emo-FilM仅作外部复核:30人观看Sintel的BOLD和同步心率,未把50项共识评分作为该复核的核心输入。
→
如何分析
对Emo-FilM的116脑区信号使用26秒滑窗、6.5秒步长,计算网络整合与分离指标,再与同期心率相关。对比主研究iEEG与fMRI,并用血流动力学模型生成模拟BOLD。
→
得到什么结果
Emo-FilM复核中,心率与分离正相关r=0.11,与整合负相关r=−0.16;方向与主研究iEEG相反、与实测fMRI一致。模拟BOLD却保留iEEG方向,说明简单血流转换并未复现实测差异。
解读边界:Emo-FilM没有提供此研究的颅内EEG。不能把心率直接等同于某一种主观情绪,也不能说模型已证明符号反转的唯一机制;当前核到的是预印本。
核对位置:Results: Replication of BOLD findings / Balloon–Windkessel modelling;Methods: External fMRI data for validation。
原始论文/作者摘要 ↗ · 本轮另核对的公开来源 ↗
05 / 脑数据的泛化与测量边界
Inter-subject correlations and their behavioral associations vary across movies
2026 · Imaging Neuroscience · 正式全文:Emo-FilM方法及跨数据集结果已核对
一部电影上发现的脑同步与行为关系,换一部电影还成立吗?
用了哪些数据
3套数据共318人、36部电影。Emo-FilM是第二复核数据:28人(排除2名缺失者)、14部电影fMRI和大五人格问卷;不是用50项连续情绪训练分类器。
→
如何分析
逐电影计算被试间脑时间序列相关(ISC);用被试间表征相似性分析(IS-RSA)考察脑同步与行为相似性的关系,再比较不同电影的空间分布。
→
得到什么结果
脑同步强度及ISC–行为关联显著依赖具体电影;不同电影的关联区域重合有限,三个数据集给出一致方向的证据。
解读边界:它是对电影fMRI结论可推广性的检验,不能解释成某一情绪项目本身可靠性差,更不能拿单片最高结果代表普遍规律。
核对位置:Methods §2.3;Introduction的三数据集设计;Results与Discussion跨电影复核。
原始论文/作者摘要 ↗ · 本轮另核对的公开来源 ↗
06 / 视觉编码与重建:不等于情绪解码
Deep Neural Encoder-Decoder Model to Relate fMRI Brain Activity with Naturalistic Stimuli
2025 · EMBC / arXiv公开稿 · 公开全文:方法、结果与讨论已核对
连续电影画面能否预测视觉皮层BOLD,并借助解码器重建画面特征?
用了哪些数据
30人的电影fMRI先跨人平均,与14部电影的图像帧配对;选取视觉区高信噪比4,609体素。不使用50项情绪评分。
→
如何分析
CNN编码器从连续32帧预测fMRI,解码器作重建与正则化。留出一整部YouAgain及其余影片末20%测试,再看显著性图中的重要体素。
→
得到什么结果
报告体素时间预测平均r≈0.203、重建SSIM≈0.320;重要区域集中在枕叶、梭状回、距状区,重建主要体现边缘、对比、脸等视觉特征。
解读边界:论文主要结果的解码器输入包含编码器预测的fMRI,不能将这些结果全部宣传成“由独立实测脑活动直接读出画面”。SSIM不是情绪识别准确率,也未检验50项目的心理结构。
核对位置:Methods §II-A–C;Results §III-A–B,尤其解码输入说明与Fig.2–3。
原始论文/作者摘要 ↗
07 / 情绪与脑活动的动态/条件差异
Neural state shifts in response to changing affective context
2026 · CNS会议海报摘要 · 仅摘要;非完整论文证据
情绪变化更稳定的电影,脑状态切换也会更少吗?
用了哪些数据
Emo-FilM的28名被试电影fMRI与连续情绪评分。摘要未详列所选项目、全部预处理与排除原因。
→
如何分析
HMM分割多体素脑活动状态;根据affective surprise的稳定程度对电影作对照,比较脑区的状态切换率。
→
得到什么结果
初步结果中,角回和中枕叶在较稳定情绪情境下切换较少;杏仁核与海马反而较多。提示不同脑区可能有不同的情绪—分段关系。
解读边界:只能作为动态加工的初步线索。摘要中的affective surprise不能未经方法核对就等同于补表Surprise单项;不能宣称已建立完整机制。
核对位置:作者海报摘要全文;未取得完整方法、数据和统计附件。
原始论文/作者摘要 ↗
08 / 情绪与脑活动的动态/条件差异
Investigating higher-order signatures of emotions in the human brain using film fMRI
2026 · ABIM会议摘要 W16 · 仅摘要;非完整论文证据
三个以上脑区的联合动态,是否比两两连接更能对应情绪时间变化?
用了哪些数据
摘要报告30人、14部电影fMRI及50项时间分辨情绪评分。
→
如何分析
计算高阶交互与同调scaffold动态,用时间—时间相似性及偏最小二乘相关(PLSC)联系脑动态和情绪;并报告交叉验证预测比较。
→
得到什么结果
作者报告高阶scaffold与情绪曲线的对应和预测优于传统两两连接,并得到与效价、唤醒、力量有关的脑—情绪成分。
解读边界:摘要未给完整性能数值、训练测试拆分和对照细节;本轮不将其结论升格为已充分复核的完整论文结论。
核对位置:ABIM官方会议册W16,Lorenzo Zaffina等。
原始论文/作者摘要 ↗
另:综述中的实际评分示例
Extending insights from LeDoux: using movies to study subjective, clinically meaningful experiences in neuroscience
该综述的Fig.1使用Sintel的50项自报评分,取前两个主成分画出情绪随时间的轨迹,说明“整部电影只贴一个情绪标签”会遗漏动态与多成分信息。这是说明性作图与方法论论证,不是一项完成情绪脑解码的新实验。综述原文 ↗
读完后,怎样准确概括它的内涵?
数据集把“情绪是什么”的理论描述落实成了多条可随电影时间变化的评分曲线。后续研究既可以检查这些描述如何组织和相互预测,也可以把它们与脑活动对齐,研究表征、状态变化或场景对个体差异的影响。结论的强度取决于所用数据和检验,不能由“多模态、50项”这些标签自动推出。