CARDIO-Affect: A Hamiltonian-Variability Framework for Spatio-Temporal Emotional Pattern Recognition with Manifold-Based Individual and Group Profiling

Xiao Sun

arxiv: 2510.16046 · v3 · pith:7D2ONV6Tnew · submitted 2025-10-16 · ⚛️ physics.soc-ph · cs.CY· cs.SI

CARDIO-Affect: A Hamiltonian-Variability Framework for Spatio-Temporal Emotional Pattern Recognition with Manifold-Based Individual and Group Profiling

Xiao Sun This is my paper

Pith reviewed 2026-05-21 21:07 UTC · model grok-4.3

classification ⚛️ physics.soc-ph cs.CYcs.SI

keywords emotional dynamicscomplex systemsHamiltonian SDEinformation geometrysocial contagiontopological data analysislongitudinal emotion data

0 comments

The pith

CARDIO-Affect models individual emotions as multi-stable nonlinear stochastic systems and group emotions as sparsely coupled networks with emergent macrostates.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper introduces CARDIO-Affect as a framework for long-term emotional dynamics within stable small groups by drawing on complex systems ideas. It represents each person's emotions as a nonlinear stochastic dynamical system that can settle into multiple stable states and represents the group's emotions as a network with only sparse connections that nonetheless produce larger collective patterns. The method combines a Hamiltonian stochastic differential equation with neural parameters, information geometry on a 45-dimensional manifold, topological trajectory analysis, and multi-scale variability measures drawn from heart-rate ideas. These tools are applied to a 30-month collection of real-world facial emotion recordings to identify three specific patterns in how emotions spread and last.

Core claim

CARDIO-Affect formalizes individual emotion as a multi-stable nonlinear stochastic dynamical system and group emotion as a sparsely-coupled network with emergent macrostates, validated by discovering Sparse-Contagion (R_0=0.36, density 2.7%), Asymmetric-Persistence (negative dwell 5.85x positive), and Crisis-Inversion paradoxes on the 30.1-month corpus.

What carries the argument

Neural-parameterised Hamiltonian SDE over asymmetric potentials, paired with information geometry on a 45-dimensional Fisher-Rao manifold and Emotional Variability Analytics that decomposes each person-day into multi-scale measures.

If this is right

Group-level emotion exhibits sparse contagion characterized by a basic reproduction number of 0.36 and only 2.7 percent edge density.
Negative emotional states persist 5.85 times longer than positive states, corresponding to a 1.77-dimensional potential gap.
Apparent emotional shifts during events such as the 2022 Shanghai lockdown become statistically insignificant once synthetic-control and permutation methods are applied.
The linear mask-self estimator matches asymptotically optimal Granger performance on linear data but underperforms on nonlinear coupled systems.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

The same multi-stable attractor picture could be used to forecast when a small group will cross from one dominant emotional regime to another.
Manifold coordinates derived for each person might serve as stable individual signatures for tracking changes in emotional resilience over years.
The sparse-coupling assumption invites direct tests in larger or online groups to see whether the low contagion density persists outside tightly bounded settings.

Load-bearing premise

The 30.1-month longitudinal in-the-wild facial-emotion corpus accurately captures naturalistic long-term emotional dynamics without substantial biases from collection methods, annotation, or environmental factors.

What would settle it

Collecting an independent multi-year facial-emotion dataset from stable groups and finding that the measured contagion density is substantially different from 2.7 percent or that the dwell-time asymmetry between negative and positive states is not near a factor of 5.85.

Figures

Figures reproduced from arXiv: 2510.16046 by Xiao Sun.

**Figure 1.** Figure 1: Emotional Contagion Network Structure. (A) Network visualization showing 592 significant Granger causality connections among 38 individuals. Node size represents out-degree (influence). Red nodes indicate emotional hubs (top 90th percentile, n = 7). Edge opacity reflects connection strength. (B) Out-degree distribution showing right-skewed pattern characteristic of scale-free networks, with a small number … view at source ↗

**Figure 2.** Figure 2: Temporal Dynamics of Emotional Contagion. (A) Temporal distribution of Granger-causality links showing that 47.8% of emotional influence occurs within 1-day lag, indicating rapid contagion. (B) Cumulative percentage of contagion events over time, demonstrating that nearly half of all emotional transmission happens within the first day. (NPI): NPI = |r(Neuroticism,Influence)| |r(Extraversion,Influence)| (7… view at source ↗

**Figure 3.** Figure 3: Personality Correlates of Emotional Influence. (A) Neuroticism shows positive correlation with emotional influence (r = 0.457, p = 0.087, marginally significant). (B) Extraversion shows no significant relationship (r = −0.066, p = 0.814). (C) Conscientiousness shows negative correlation (r = −0.077, p = 0.784). Scatter plots with regression lines demonstrate that emotional instability, not extraversion, pr… view at source ↗

**Figure 4.** Figure 4: Temporal Dynamics of Emotional Contagion. (A) Distribution of optimal Granger lags showing 47.8% of connections manifest within 1 day. (B) Cumulative distribution function of transmission times. (C) Comparison of transmission speed between hubs and non-hubs. Hubs transmit significantly faster (mean lag=2.54 days vs. 3.18 days, t = −3.076, p = 0.0022). (D) Reciprocity analysis showing 70.3% of connections a… view at source ↗

**Figure 5.** Figure 5: Entropy Dynamics and Temporal Evolution. (A) Emotional variance over time showing 22.9% increase from first half to second half of study period. (B) Individual-level variance changes: 31 of 38 individuals (81.6%) showed increased variance. (C) Autocorrelation decay analysis showing emotions return to baseline within 3–5 days, but baseline variance itself increases. (D) Hub vs. non-hub variance trajectories… view at source ↗

**Figure 6.** Figure 6: Affective Epidemiology Framework. Integrated model showing how network structure, personality traits, and temporal dynamics interact to produce collective emotional patterns. (A) Conceptual diagram of the Neuroticism Paradox mechanism. (B) Mathematical formulation of emotional contagion as an epidemic process. (C) Predictive model showing how hub removal would affect system dynamics. (D) Comparison with t… view at source ↗

**Figure 7.** Figure 7: Extended Network Analysis. (A) Degree distribution of the emotional contagion network showing right-skewed distribution characteristic of scale-free networks. (B) Betweenness centrality distribution identifying key bridge nodes. (C) Comparison of hub characteristics across different centrality measures. (D) Temporal evolution of network density over the 30.5-month period. Data Availability Anonymized data … view at source ↗

**Figure 8.** Figure 8: Statistical Validation and Robustness Checks. (A) Permutation test results showing observed network density far exceeds random expectations. (B) Phase randomization surrogate analysis confirming genuine emotional coupling. (C) Sensitivity analysis of R0 to Granger p-value threshold. (D) Bootstrap confidence intervals for personality-influence correlations. References 1. Hatfield, E., Cacioppo, J. T., & Ra… view at source ↗

**Figure 9.** Figure 9: Verified Emotional Contagion Network (15 High-Quality Participants). Network visualization of emotional influence among 15 participants with ≥100 days of data. Node size represents emotional influence (out-degree), node color represents neuroticism level (red = high, blue = low). The figure demonstrates that high-neuroticism individuals (red nodes) occupy central positions with larger node sizes, confirmi… view at source ↗

**Figure 10.** Figure 10: Temporal Evolution of Emotional Variance. (A) System-wide emotional variance over the 30.5-month period showing overall trend. (B) Variance trajectories separated by neuroticism level (High Neuroticism n=3, Low Neuroticism n=3). High-neuroticism individuals show consistently higher variance and greater volatility, supporting the mechanism that emotionally unstable individuals drive system-wide entropy i… view at source ↗

read the original abstract

We present CARDIO-Affect, a complex-systems theoretical framework for long-term emotional dynamics in bounded social groups, with explicit uncertainty quantification at every layer. Long-period naturalistic emotion in stable small groups exhibits hallmarks of complex systems -- multi-stable attractors, weak chaos, long-range memory, and sparse heterogeneous coupling -- invisible to conventional short-clip facial-emotion analysis. CARDIO-Affect treats individual emotion as a multi-stable nonlinear stochastic dynamical system and group emotion as a sparsely-coupled network with emergent macrostates, formalised through six propositions and four pillars: (i) statistical mechanics with neural-parameterised Hamiltonian SDE over asymmetric potentials; (ii) information geometry on a 45-dimensional Fisher-Rao manifold; (iii) topological data analysis for invariant trajectory signatures; (iv) HRV-inspired Emotional Variability Analytics (EVA) decomposing each person-day into multi-scale time/frequency/nonlinear measures. We validate on the first 30.1-month longitudinal in-the-wild facial-emotion corpus (companion: arXiv:2510.15221) by discovering three falsifiable paradoxes: Sparse-Contagion (R_0=0.36, density 2.7%, 8 BH-FDR edges), Asymmetric-Persistence (negative dwell 5.85x positive, 1.77D potential gap), and Crisis-Inversion (Shanghai 2022 lockdown naive d=-0.40 collapses to permutation-p=0.94 under BSTS + synthetic-control). On synthetic benchmarks, CARDIO-EBM v2 matches asymptotically optimal Granger on linear VAR data (Class A AUROC 0.984+/-0.012 vs Granger 0.997+/-0.001, 5 seeds) but fails on tanh-coupled nonlinear data (Class B AUROC 0.490 vs Granger 0.796), a documented limitation of the linear mask-self estimator. We release framework code and the full reproduction pipeline.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

CARDIO-Affect offers a fresh integration of Hamiltonian SDEs, Fisher-Rao geometry, and variability metrics for long-term group emotion modeling, but the validation on the 30-month corpus rests on thin evidence with clear gaps in controls and derivations.

read the letter

This paper introduces CARDIO-Affect as a framework that treats emotions in small groups as complex dynamical systems with multi-stable attractors and sparse coupling. It uses a neural-parameterized Hamiltonian stochastic differential equation on a 45-dimensional Fisher-Rao manifold, plus topological data analysis and variability metrics inspired by heart rate variability. What is new is the specific integration of these tools into one pipeline for spatio-temporal profiling over months rather than short clips. The discovery of three paradoxes—Sparse-Contagion, Asymmetric-Persistence, and Crisis-Inversion—on their 30-month corpus is presented as validation. The paper does well by releasing the code and full reproduction pipeline, which allows others to check the work. The synthetic benchmarks provide some external check, even if they highlight limitations. The soft spots are more concerning. The method performs well on linear synthetic data but fails on nonlinear cases, which seems at odds with the complex systems claims. There are no visible error bars, full derivation steps, or details on how the facial data was cleaned or controlled for confounds like lighting and pose. This leaves room for the paradoxes to be artifacts from the data collection rather than genuine dynamics, as the stress test suggests. The circularity in fitting and validating on the same corpus adds to the caution. This work is for people in affective computing or complex systems approaches to psychology who want to move beyond snapshot analyses. A reader looking for formal models of group emotion might get ideas from it, though they would need to dig into the methods. It deserves a serious referee because the formalization is ambitious and the ideas could spark useful discussion, even with the current gaps. I recommend sending it to peer review for a closer look at the derivations and data handling.

Referee Report

3 major / 2 minor

Summary. The paper presents CARDIO-Affect, a complex-systems framework that models individual emotion as a multi-stable nonlinear stochastic dynamical system via neural-parameterised Hamiltonian SDE over asymmetric potentials and group emotion as a sparsely-coupled network with emergent macrostates. It formalises this through six propositions and four pillars (statistical mechanics, 45-dimensional Fisher-Rao information geometry, topological data analysis, and HRV-inspired Emotional Variability Analytics), validates the approach on a 30.1-month in-the-wild facial-emotion corpus by discovering three paradoxes (Sparse-Contagion with R_0=0.36 and 2.7% density, Asymmetric-Persistence with 5.85x negative dwell asymmetry, and Crisis-Inversion), reports synthetic benchmarks, and releases code plus reproduction pipeline.

Significance. If the results hold, the work would advance socio-physics and affective computing by providing a Hamiltonian-variability lens for long-term emotional dynamics with explicit uncertainty quantification and manifold-based profiling. The release of framework code and the full reproduction pipeline is a clear strength supporting reproducibility and falsifiability.

major comments (3)

[Abstract] Abstract, synthetic benchmarks paragraph: the documented failure of CARDIO-EBM v2 on tanh-coupled nonlinear data (Class B AUROC 0.490 vs Granger 0.796) directly undermines support for the central claim that the framework captures multi-stable nonlinear stochastic dynamical systems, as this limitation is load-bearing for the six propositions.
[Validation section] Validation section: dwell times, potential gaps, and contagion metrics (including R_0=0.36 and the three paradoxes) are derived from models fitted to the same 30.1-month longitudinal corpus used for validation, creating a circularity risk that is not mitigated by explicit separation of fitting and testing procedures or by the post-hoc BSTS + synthetic-control controls.
[Abstract] Abstract: quantitative results such as R_0=0.36, AUROC 0.984, and the 5.85x dwell asymmetry lack visible error bars, full derivation steps, or data exclusion details, which weakens the evidential link between the corpus and the claimed paradoxes.

minor comments (2)

[Abstract] The abstract references the companion data paper (arXiv:2510.15221) but does not briefly summarise how it addresses potential collection, annotation, or environmental biases in the 30.1-month corpus.
[Methods] Notation for the 45-dimensional Fisher-Rao manifold and its embedding within the Hamiltonian SDE could be clarified with an explicit coordinate definition or diagram in the methods.

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for their constructive and detailed comments. We address each major comment point by point below, with honest assessment of where revisions are warranted and where we maintain the original framing while improving clarity.

read point-by-point responses

Referee: [Abstract] Abstract, synthetic benchmarks paragraph: the documented failure of CARDIO-EBM v2 on tanh-coupled nonlinear data (Class B AUROC 0.490 vs Granger 0.796) directly undermines support for the central claim that the framework captures multi-stable nonlinear stochastic dynamical systems, as this limitation is load-bearing for the six propositions.

Authors: The benchmark paragraph explicitly documents the limitation of the linear mask-self estimator on tanh-coupled nonlinear data, and we agree this scope restriction should be stated more prominently. However, the central claim that the framework captures multi-stable nonlinear stochastic dynamical systems rests on the neural-parameterised Hamiltonian SDE over asymmetric potentials (pillar i), the 45D Fisher-Rao geometry, TDA invariants, and EVA decomposition, not solely on the contagion estimator. The six propositions are formalised at the level of the full Hamiltonian-variability framework and are supported by the three paradoxes discovered on the longitudinal corpus. We will revise the abstract to separate the estimator-specific benchmark result from the broader framework claims and to note the documented limitation of the linear mask-self component. revision: partial
Referee: [Validation section] Validation section: dwell times, potential gaps, and contagion metrics (including R_0=0.36 and the three paradoxes) are derived from models fitted to the same 30.1-month longitudinal corpus used for validation, creating a circularity risk that is not mitigated by explicit separation of fitting and testing procedures or by the post-hoc BSTS + synthetic-control controls.

Authors: We acknowledge that the metrics are derived from parameters estimated on the full corpus, which introduces a potential circularity concern in a discovery-oriented longitudinal study. The BSTS + synthetic-control analysis is applied specifically to the Crisis-Inversion paradox to provide a counterfactual check. For the remaining metrics we will add an explicit subsection describing the fitting procedure, any robustness or sensitivity checks performed, and the logical separation between parameter estimation (to recover the dynamical system) and the subsequent computation of dwell times, potential gaps, and contagion statistics. We will also discuss why full hold-out testing is not straightforward in this 30-month in-the-wild setting while strengthening the description of post-fitting validation steps. revision: yes
Referee: [Abstract] Abstract: quantitative results such as R_0=0.36, AUROC 0.984, and the 5.85x dwell asymmetry lack visible error bars, full derivation steps, or data exclusion details, which weakens the evidential link between the corpus and the claimed paradoxes.

Authors: We agree that the absence of error bars, derivation outlines, and explicit data-exclusion criteria weakens the presentation of the quantitative results. In the revised manuscript we will (i) report confidence intervals or standard deviations for R_0, the AUROC values (already shown with ±0.012 across seeds in the full text), and the dwell asymmetry ratio, (ii) add a concise derivation summary for the dwell-time and potential-gap calculations in the methods or supplementary material, and (iii) state the data-exclusion criteria applied to the 30.1-month corpus. These additions will be reflected in both the abstract and the main validation section. revision: yes

Circularity Check

1 steps flagged

Dwell times, potential gaps and contagion metrics extracted from models fitted to the validation corpus

specific steps

fitted input called prediction [Abstract]
"We validate on the first 30.1-month longitudinal in-the-wild facial-emotion corpus (companion: arXiv:2510.15221) by discovering three falsifiable paradoxes: Sparse-Contagion (R_0=0.36, density 2.7%, 8 BH-FDR edges), Asymmetric-Persistence (negative dwell 5.85x positive, 1.77D potential gap), and Crisis-Inversion (Shanghai 2022 lockdown naive d=-0.40 collapses to permutation-p=0.94 under BSTS + synthetic-control)."

The reported dwell times, potential gap and contagion density are computed from the Hamiltonian SDE and sparsely-coupled network that were fitted to the identical 30.1-month corpus; the 'discovery' of these quantitative signatures is therefore a direct statistical consequence of the model fit rather than an independent test of the framework.

full rationale

The framework is formalized via six propositions and four pillars (Hamiltonian SDE, Fisher-Rao manifold, TDA, EVA) that are independent of the corpus. Synthetic benchmarks on linear VAR and tanh-coupled data are also external. However, the load-bearing validation step discovers the three paradoxes directly from the same 30.1-month corpus to which the multi-stable SDE and sparse network are fitted; dwell times, potential gaps and R_0 are therefore outputs of that fit. This produces one clear instance of fitted-input-called-prediction, yielding partial circularity (score 6) while leaving the theoretical core and synthetic checks intact.

Axiom & Free-Parameter Ledger

2 free parameters · 1 axioms · 1 invented entities

Ledger inferred from abstract; full paper would list additional fitted values and background results. Many modeling choices appear ad hoc to enable the complex-systems treatment of emotion.

free parameters (2)

45-dimensional Fisher-Rao manifold
Dimension chosen for information geometry pillar to represent emotional state distributions.
R_0=0.36 and density 2.7%
Effective reproduction number and edge density fitted from group interaction data for Sparse-Contagion paradox.

axioms (1)

domain assumption Long-period naturalistic emotion in stable small groups exhibits multi-stable attractors, weak chaos, long-range memory, and sparse heterogeneous coupling.
Stated as the foundational observation enabling the complex-systems treatment in the framework introduction.

invented entities (1)

Emotional Variability Analytics (EVA) no independent evidence
purpose: Decompose each person-day into multi-scale time/frequency/nonlinear measures for individual profiling.
New analytic component introduced as the fourth pillar, inspired by but distinct from HRV.

pith-pipeline@v0.9.0 · 5899 in / 1630 out tokens · 72711 ms · 2026-05-21T21:07:12.564075+00:00 · methodology

discussion (0)

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

IndisputableMonolith/Cost/FunctionalEquation.lean washburn_uniqueness_aczel unclear

?

unclear
Relation between the paper passage and the cited Recognition theorem.

CARDIO-Affect treats individual emotion as a multi-stable nonlinear stochastic dynamical system ... formalised through ... (i) statistical mechanics with neural-parameterised Hamiltonian SDE over asymmetric potentials
IndisputableMonolith/Foundation/AlexanderDuality.lean alexander_duality_circle_linking unclear

?

unclear
Relation between the paper passage and the cited Recognition theorem.

information geometry on a 45-dimensional Fisher-Rao manifold; topological data analysis for invariant trajectory signatures

What do these tags mean?

matches: The paper's claim is directly supported by a theorem in the formal canon.
supports: The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends: The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses: The paper appears to rely on the theorem as machinery.
contradicts: The paper's claim conflicts with a theorem or certificate in the canon.
unclear: Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

12 extracted references · 12 canonical work pages

[1]

T., & Rapson, R

Hatfield, E., Cacioppo, J. T., & Rapson, R. L. (1993). Emotional contagion.Current Direc- tions in Psychological Science, 2(3), 96–100

work page 1993
[2]

Barsade, S. G. (2002). The ripple effect: Emotional contagion and its influence on group behavior.Administrative Science Quarterly, 47(4), 644–675

work page 2002
[3]

A., & Fowler, J

Christakis, N. A., & Fowler, J. H. (2009).Connected: The surprising power of our social networks and how they shape our lives. Little, Brown

work page 2009
[4]

Goldenberg, A., & Gross, J. J. (2020). Digital emotion contagion.Trends in Cognitive Sciences, 24(4), 316–328

work page 2020
[5]

Granger, C. W. J. (1969). Investigating causal relations by econometric models and cross- spectral methods.Econometrica, 37(3), 424–438

work page 1969
[6]

M., & May, R

Anderson, R. M., & May, R. M. (1991).Infectious diseases of humans: Dynamics and control. Oxford University Press

work page 1991
[7]

J., & Strogatz, S

Watts, D. J., & Strogatz, S. H. (1998). Collective dynamics of ‘small-world’ networks.Nature, 393(6684), 440–442

work page 1998
[8]

Schilbach, L., et al. (2013). Toward a second-person neuroscience.Behavioral and Brain Sciences, 36(4), 393–414

work page 2013
[9]

Singer, T., & Lamm, C. (2009). The social neuroscience of empathy.Annals of the New York Academy of Sciences, 1156(1), 81–96. 21

work page 2009
[10]

Herrando, C., & Constantinides, E. (2021). Emotional contagion: A brief overview and future directions.Frontiers in Psychology, 12, 712606

work page 2021
[11]

D., Guillory, J

Kramer, A. D., Guillory, J. E., & Hancock, J. T. (2014). Experimental evidence of massive- scale emotional contagion through social networks.Proceedings of the National Academy of Sciences, 111(24), 8788–8790

work page 2014
[12]

A., & LeDoux, J

Phelps, E. A., & LeDoux, J. E. (2005). Contributions of the amygdala to emotion processing: From animal models to human behavior.Neuron, 48(2), 175–187. Acknowledgments This work was supported by Special Project of the National Natural Science Foundation of China (62441614), Anhui Province Key R&D Program (202304a05020068) and General Programmer of the Na...

work page 2005

[1] [1]

T., & Rapson, R

Hatfield, E., Cacioppo, J. T., & Rapson, R. L. (1993). Emotional contagion.Current Direc- tions in Psychological Science, 2(3), 96–100

work page 1993

[2] [2]

Barsade, S. G. (2002). The ripple effect: Emotional contagion and its influence on group behavior.Administrative Science Quarterly, 47(4), 644–675

work page 2002

[3] [3]

A., & Fowler, J

Christakis, N. A., & Fowler, J. H. (2009).Connected: The surprising power of our social networks and how they shape our lives. Little, Brown

work page 2009

[4] [4]

Goldenberg, A., & Gross, J. J. (2020). Digital emotion contagion.Trends in Cognitive Sciences, 24(4), 316–328

work page 2020

[5] [5]

Granger, C. W. J. (1969). Investigating causal relations by econometric models and cross- spectral methods.Econometrica, 37(3), 424–438

work page 1969

[6] [6]

M., & May, R

Anderson, R. M., & May, R. M. (1991).Infectious diseases of humans: Dynamics and control. Oxford University Press

work page 1991

[7] [7]

J., & Strogatz, S

Watts, D. J., & Strogatz, S. H. (1998). Collective dynamics of ‘small-world’ networks.Nature, 393(6684), 440–442

work page 1998

[8] [8]

Schilbach, L., et al. (2013). Toward a second-person neuroscience.Behavioral and Brain Sciences, 36(4), 393–414

work page 2013

[9] [9]

Singer, T., & Lamm, C. (2009). The social neuroscience of empathy.Annals of the New York Academy of Sciences, 1156(1), 81–96. 21

work page 2009

[10] [10]

Herrando, C., & Constantinides, E. (2021). Emotional contagion: A brief overview and future directions.Frontiers in Psychology, 12, 712606

work page 2021

[11] [11]

D., Guillory, J

Kramer, A. D., Guillory, J. E., & Hancock, J. T. (2014). Experimental evidence of massive- scale emotional contagion through social networks.Proceedings of the National Academy of Sciences, 111(24), 8788–8790

work page 2014

[12] [12]

A., & LeDoux, J

Phelps, E. A., & LeDoux, J. E. (2005). Contributions of the amygdala to emotion processing: From animal models to human behavior.Neuron, 48(2), 175–187. Acknowledgments This work was supported by Special Project of the National Natural Science Foundation of China (62441614), Anhui Province Key R&D Program (202304a05020068) and General Programmer of the Na...

work page 2005