An Adversarial Audit of TSTOEAO’s Route-Selection Architecture; A Proponent-Run Adversarial Audit

DOI: To be assigned

John Swygert

July 26, 2026

A TSTOEAO adversarial evidence paper

Abstract

A universal theory must do more than recognize patterns; it must reject false matches, distinguish levels of evidence, and state what would count against it. This paper performs a bounded adversarial audit of twenty-one external studies and claims previously considered for relevance to the Swygert Theory of Everything AO (TSTOEAO). The audit applies locked eligibility gates and a four-axis scoring system covering source status, chronology, mechanistic specificity, and discriminating controls. The evidence universe includes supportive experiments, theoretical work, review articles, secondary reporting, company claims, and studies that predate the relevant TSTOEAO propositions. The audit rejects or downgrades cases when the primary work predates the theory, when only a review or news article is available, when the mechanistic mapping is merely verbal, when no identifiable change in relational architecture occurs, or when independent data are insufficient. One case qualifies as direct domain-specific confirmation; six qualify as strong post-publication corroboration; five receive qualified corroborative or conceptual status; five are retained only as conventional baseline; three are rejected as confirmation because they are secondary or insufficiently grounded; and one is retained as review-level context. No direct empirical contradiction is found in this bounded corpus, which limits the strength of the audit and motivates a prospective program of deliberately selected null and control cases. The result is not that TSTOEAO has already passed every possible adversarial test. The result is that the framework can state clear exclusion rules, apply them against its own favored examples, and convert its central relation V = E x Y from an open interpretive slogan into a route-selection architecture with specific failure conditions.

This paper demonstrates evidentiary exclusion and specifies empirical failure conditions; it does not claim that a direct TSTOEAO experimental falsification has already been observed.

Keywords: TSTOEAO; route selection; adversarial audit; falsifiability; Encoded Equilibrium; evidence classification; negative controls; preregistration; receiver dependence; cost-location

1. Introduction: A Universal Theory Must Exclude

The strongest criticism of a universal theory is not that it sees a pattern. It is that it may see the pattern everywhere. A framework that can explain every possible result after the fact has no meaningful edge. To become scientifically useful, it must distinguish a true instance of its architecture from a superficial resemblance, a pre-existing conventional result, a secondary article, an unverified claim, and an observation that would count against it.

The preceding paper, “The Prediction Is the Pattern,” argued that the central TSTOEAO prediction is architectural: across different domains, Energy or Opportunity becomes measurable Value through Encoded Equilibrium, the relational structure that determines available routes. It formalized Y as a route-selection operator and identified recurring components of boundary, geometry, phase, connectivity, interface, frequency, receiver, and cost-location.

V = E x Y

That argument established a cumulative confirmation case. It did not by itself establish that the evidence review was immune to selection bias. The present paper addresses that weakness directly. It asks whether TSTOEAO can say no to material that appears attractive, whether it can downgrade a correspondence without discarding it, and whether it can state observations that would weaken the operational law.

The objective is not to manufacture a contradiction. It is to expose the decision boundary. A theory that says no only when a paper is inconvenient is not disciplined. A theory that states its rejection rules in advance and applies them to its own favored examples has begun to become auditable.

2. The Claim Under Audit

The audited claim is narrower than the full ontology of TSTOEAO. It concerns the operational route-selection law:

Comparable Energy or Opportunity can produce different measurable Values because Encoded Equilibrium changes the admissible routes, their relative weights, the receiver-accessible record, and the location at which correction is paid.

In operator form, let A(Y) denote the routes permitted by the current relational architecture. A system’s observable result can be represented schematically as:

V = M [ sum over r in A(Y) of w_r(E,Y) T_r(E) ]

where T_r is a route transformation, w_r is its accessibility or weight, and M is the receiver or measurement operation. The expression is not offered as a replacement for domain equations. It specifies the common causal grammar that a qualifying case must display.

A paper does not qualify merely because it reports a structure-property relationship. A strong case requires an identifiable change in Y, a measurable change in V, and a causal account of how the intervention selected, suppressed, redirected, or reconstructed a route.

3. Evidence Universe and Lock Point

This audit uses a bounded evidence universe rather than claiming to sample all scientific literature. The universe consists of twenty-one external studies, programs, reviews, secondary articles, or claims that were explicitly considered for TSTOEAO relevance in the working research sequence preceding this paper. The set includes the ten principal examples used in the prior convergence paper, additional later candidates, and cases previously downgraded or excluded.

The lock point for this paper occurs before the case-by-case results are presented: the eligibility gates, scoring axes, and decision rules below are fixed first and then applied to all twenty-one items. This is not a public preregistration and should not be represented as one. It is a transparent, auditable proponent-run adversarial reassessment. Future audits should be time-stamped before the evidence window closes and should use independent scorers.

The evidence universe is deliberately heterogeneous. That is necessary because an honest filter must be able to reject not only irrelevant experiments, but also attractive secondary stories, reviews, company announcements, and strong mechanisms that fail the chronology requirement for later confirmation.

4. Locked Audit Method

4.1 Eligibility gates

1. Primary-source gate: a claim cannot qualify as empirical confirmation when only a secondary article, press account, or company announcement is available.
2. Chronology gate: work publicly available before the relevant TSTOEAO proposition cannot count as later confirmation, regardless of mechanistic fit.
3. Intervention gate: the case must identify what changed in boundary, geometry, phase, connectivity, interface, frequency, receiver, or cost-location.
4. Observable gate: the case must identify a measured or formally derived V rather than a general promise of future capability.
5. Mechanism gate: the chain from intervention to route change to observable must be explicit enough to audit.
6. Scope gate: a local result cannot be promoted into proof of the entire TSTOEAO ontology.

4.2 Four-axis score

Each case receives four scores. The numbers do not constitute a Bayesian probability or a universal metric. They make the classification decisions visible and reproducible.

• S – Source and empirical status, 0-2: 0 secondary or company claim; 1 theory, simulation, review, or program-level evidence; 2 primary experiment.
• C – Public chronology, 0-2: 0 predates the relevant proposition; 1 publicly later than the foundation but not a fulfilled domain-specific proposition; 2 publicly later than a specific prior TSTOEAO domain proposition.
• M – Mechanistic specificity, 0-3: 0 no usable mapping; 1 general structure-property relevance; 2 explicit Y intervention and V change; 3 route selection, portfolio reallocation, receiver dependence, cost-location, or record recovery.
• D – Discrimination and controls, 0-3: 0 none or unavailable; 1 one relevant comparison; 2 multiple controls, threshold tests, orientation tests, component ablations, or matched contrasts; 3 a TSTOEAO-specific prospective test locked before the outcome.

Audit score Q = S + C + M + D

A total score does not override the gates. A pre-2025 paper may score highly on mechanism and controls while remaining ineligible as later confirmation. A secondary article cannot be promoted by a high conceptual resemblance. No case in this retrospective corpus receives D = 3, because none was a TSTOEAO-specific prospective test locked before its data were known. Q is a transparent descriptive rubric, not a validated measurement scale. The classifications do not depend upon a rigid distinction between adjacent totals such as Q = 7 and Q = 8; eligibility gates, source quality, chronology, mechanism, and evidentiary limits govern the final classification.

4.3 Decision classes

• Direct domain-specific confirmation: later primary experiment corresponding to a specific prior domain proposition, with explicit intervention, observable, and controls.
• Strong post-publication corroboration: later primary experiment with explicit route architecture and strong controls, but no prior numerical or local point prediction.
• Qualified corroboration or theoretical convergence: relevant and mechanically meaningful, but limited by source status, incomplete route accounting, or non-unique interpretation.
• Conventional baseline: strong relevant science that predates TSTOEAO and therefore cannot be claimed as later confirmation.
• Review-level context: synthesis useful for organization but not a new empirical event.
• Unresolved: potentially relevant but insufficiently independently supported.
• Reject as confirmation: secondary, merely thematic, chronologically ineligible, or lacking a causal intervention-observable chain.

4.4 Automatic no conditions

• No primary data, no empirical confirmation.
• No later public chronology, no post-publication confirmation.
• No identifiable Y change, no route-selection claim.
• No measurable V, no completed test.
• No causal chain, no mechanistic promotion.
• No distinction between equivalent frequency and measured resonance, no frequency claim.
• No measured cost channel, no claim that cost relocation was demonstrated.
• No receiver comparison, no claim that receiver dependence was tested.
• No complete ontology inference from one local experiment.

5. Audit Results

The twenty-one cases separate into a mixed record rather than a perfect batting average. One qualifies as direct domain-specific confirmation. Six qualify as strong corroboration. Five qualify only as qualified corroboration or theoretical convergence. Five are retained as conventional baseline because their public records predate TSTOEAO. One is review-level context. Two secondary articles are rejected as confirmation. One commercial claim remains unresolved.

No direct empirical contradiction is identified in this bounded corpus. That absence cannot be treated as a universal survival test because the corpus was assembled from candidates initially noticed for possible relevance. The audit therefore demonstrates exclusion discipline and evidentiary downgrading, but it does not replace a future search designed specifically to locate null and contradictory cases.

Table 1. Locked adversarial audit of the twenty-one-case evidence universe

ID

Candidate

S

C

M

D

Q

Decision and reason

1

Blue perovskite LEDs

2

2

3

2

9

Direct confirmation: Later experiment directly implements the earlier Chromatic Determinism variables.

2

Quantum statistical plasmonic metacrystals

2

1

3

2

8

Strong corroboration: Geometry creates experimentally distinct allowed and forbidden statistical bands.

3

CeTe3 competing electronic phases

2

1

3

2

8

Strong corroboration: A modest field reallocates competing measured charge and spin states.

4

Room-temperature phonon focusing

2

1

3

2

8

Strong corroboration: Crystal orientation selects different room-temperature heat-propagation symmetries.

5

Iodine/MXene Bi4Ti3O12 H2O2 piezocatalysis

2

1

3

2

8

Strong corroboration: Doping plus an electron-sink interface redirects carriers into H2O2 production.

6

Co-doped BiFeO3/Pt H2 piezocatalysis

2

1

3

2

8

Strong corroboration: Bulk polarization and Pt surface kinetics jointly select the H2 route.

7

Reconstructed-mica oxide epitaxy

2

1

2

2

7

Qualified corroboration: A reconstructed oxygen plane selects epitaxial orientation; broader costs were not isolated.

8

Frozen CS2 optoacoustic fiber

2

1

2

1

6

Qualified corroboration: A reversible phase change raises Brillouin gain; the full route portfolio was not measured.

9

Inverse effective field theory

1

1

3

1

6

Theoretical convergence: Strong record-recovery mathematics, but no new hidden particle spectrum was measured.

10

Nonplanar Josephson gauge circuit

2

1

2

2

7

Qualified corroboration: Nonplanar connectivity produces tunable gauge behavior without a prior local prediction.

11

Alkali-modified MOF glasses

2

1

2

2

7

Qualified corroboration: Modifier concentration changes network connectivity and properties through graded controls.

12

Programmable thermal nonreciprocity

2

1

3

2

8

Strong corroboration: Phase-change and magneto-optical architecture separately route absorption and emission.

13

Bilayer-nickelate leading-edge gap

2

0

3

2

7

Conventional baseline: Public arXiv preprint dated July 10, 2025; ineligible as later confirmation.

14

Au25 ligand-controlled spin-polarized emission

2

0

3

2

7

Conventional baseline: Primary article was public in May 2025; retained only as baseline.

15

High-Luminosity LHC magnet architecture

1

0

2

1

4

Conventional baseline: The core engineering program and magnet concepts long predate TSTOEAO.

16

Altermagnetism

2

0

3

2

7

Conventional baseline: Theory and direct evidence were public by 2022-2024.

17

Relativity of spacetime superpositions

1

0

2

1

4

Conventional baseline: Relevant relational framework was public in 2023.

18

Electrochemical cross-electrophile coupling review

1

1

2

0

4

Review-level context: A review organizes mixed-age studies; it is not one new experiment.

19

Language-is-not-thought secondary article

0

0

2

0

2

Reject as confirmation: Secondary commentary based mainly on pre-existing neuroscience.

20

Feelings/consciousness secondary article

0

0

2

0

2

Reject as confirmation: Secondary reporting based on older feeling and interoception literature.

21

Commercial diamond quantum processors

0

1

2

0

3

Unresolved: Relevant company claims lack sufficient independent benchmarking.

S = source/empirical status; C = public chronology; M = mechanistic specificity; D = discrimination/controls; Q = total. Eligibility gates override totals.

6. What the Theory Said No To

The most important result is not the number of accepted cases. It is the set of promotions that were refused.

7. The bilayer-nickelate result was not counted as later confirmation because its arXiv preprint was public on July 10, 2025.
8. The Au25 ligand result was not counted because the primary article was published online in May 2025.
9. The High-Luminosity LHC magnet program was not counted because its design and development history substantially predates TSTOEAO.
10. Altermagnetism was not counted because the class was formulated in 2022 and supported experimentally by 2024.
11. The spacetime-superposition perspective was not counted because the relevant public work appeared in 2023.
12. The electrochemical cross-electrophile article was not treated as one new experiment because it is a review spanning mixed chronologies.
13. The language-and-thought and feelings-and-consciousness magazine articles were not promoted into confirmation because the underlying neuroscience predates TSTOEAO and the articles are secondary sources.
14. Commercial diamond-processor claims were not promoted because independent benchmark evidence was not sufficient in the reviewed material.
15. The inverse-effective-field-theory paper was not called empirical discovery of hidden particles because it establishes a mathematical reconstruction method rather than reporting a newly reconstructed spectrum.
16. The frozen-fiber and reconstructed-mica studies were not promoted to direct confirmation because the prior TSTOEAO proposition was architectural rather than a locked local prediction of their measured values.

These refusals matter because several excluded cases are excellent science and fit the route grammar extremely well. The reason for exclusion is not lack of relevance. It is that relevance, chronology, source quality, and confirmation are different claims.

7. Discriminating Controls Within the Accepted Experiments

Although this corpus contains no direct TSTOEAO contradiction, the stronger experiments include internal contrasts that demonstrate the architecture can distinguish enabled and disabled routes.

• Blue perovskite LEDs: the complete isomeric hydrogen-bond network outperformed partial and control architectures, linking the measured gain to the combined interface and bulk design.
• Plasmonic metacrystals: fields within allowed statistical bands propagated differently from fields within forbidden bands; geometry did not merely increase all outputs.
• CeTe3: the response reorganized near a modest magnetic threshold and displayed anti-correlated competition among distinct charge-order routes.
• Phonon focusing: different crystallographic orientations produced different propagation symmetries across multiple samples rather than one universal heat pattern.
• H2O2 piezocatalysis: pristine BTO, iodine-only, MXene-only, and combined architectures produced distinct yields, isolating complementary bulk and interface contributions.
• Hydrogen piezocatalysis: cobalt-induced polarization and platinum surface sites performed distinct roles; neither raw vibration nor one component alone explained the optimized route.
• MOF glasses: modifier concentration produced systematic connectivity and transition-temperature changes, and extraction changed porosity in the predicted direction.
• Programmable thermal nonreciprocity: phase state and propagation direction produced different absorption-emission behavior, with state retention after power removal.

These controls are not TSTOEAO-specific preregistered tests. They are nevertheless stronger than uncontrolled resemblance because they show that the reported output depends upon the identified relational architecture rather than merely co-occurring with it.

8. Distinctiveness Without Inventing Rival Universal Theories

Thermodynamics, control theory, boundary-value physics, catalysis, network theory, and domain-specific material models are not rival theories of everything in the same sense as TSTOEAO. They are established local or partial frameworks. The appropriate comparison is therefore not a contest in which one must be declared universally false. The question is whether TSTOEAO adds a transferable law that preserves those local mechanisms while connecting them.

The audit identifies four possible additions beyond the statement that structure matters:

17. Route portfolio: outputs are treated as competing or coupled possibilities whose weights are redistributed by Y.
18. Cost-location: improvement in one route should be accompanied by displaced burden, dissipation, instability, latency, noise, depletion, or another measurable cost when the system is closed sufficiently for accounting.
19. Receiver dependence: the accessible record depends upon the measurement and translation architecture, not only the source process.
20. Record recoverability: observable outputs may preserve enough constraint to reconstruct features of hidden architecture.

These additions are not automatically unique. Thermodynamics already handles dissipation, control theory handles cost functions and observability, and inverse problems recover hidden parameters from measurements. TSTOEAO’s proposed contribution is their integration into one route-selection grammar applied across domains. That unification must earn its value through compression, transfer, and new tests rather than by denying the prior sciences from which its components can be recognized.

The correct claim is therefore neither that no other science explains any component nor that another complete unification already duplicates TSTOEAO. The defensible claim is that no compared local framework, by itself, supplies the entire cross-domain sequence:

gradient -> relational boundary -> route portfolio -> correction -> cost-location -> receiver record -> equilibrium target

9. Direct Failure Conditions for the Operational Law

The following observations would weaken or falsify specified forms of the route-selection law when tested under controlled conditions.

21. Y-independence: substantial controlled changes in the identified relational architecture repeatedly leave the predicted route weights and observable V unchanged.
22. Energy-only sufficiency: V is fully determined by E across conditions in which TSTOEAO predicts decisive changes from boundary, phase, geometry, interface, or receiver.
23. No portfolio coupling: the theory predicts anti-correlated reallocation among routes, but all candidate routes vary independently or in the same direction after confounds are removed.
24. No cost-location: a closed and sufficiently measured system shows durable route improvement with no displaced cost, dissipation, resource draw, instability, or uncertainty burden in any relevant channel.
25. Receiver invariance: two receivers predicted to expose different records yield indistinguishable accessible information after sensitivity and bandwidth are matched.
26. No record recoverability: observables claimed to encode hidden architecture cannot support blind reconstruction better than suitable null and conventional models.
27. Frequency failure: a predeclared measured resonance or spectral signature is absent within the stated uncertainty and bandwidth, or appears equally in matched controls.
28. Cross-domain non-transfer: the operator variables must be redefined so radically in each field that no stable intervention grammar remains.
29. Unbounded scoring: independent evaluators apply the locked audit rules and cannot distinguish high-value cases from irrelevant or deliberately mismatched controls.
30. Prospective ordering failure: a locked TSTOEAO experiment predicts an ordering among route observables and the reproducible ordering is absent or reversed.

A failure of one subprediction does not logically erase every other part of the corpus. It does require revision of the failed claim. A universal theory becomes stronger by dividing itself into propositions that can fail separately rather than defending every sentence as one indivisible object.

10. Locked Prospective Program

The next evidentiary stage should deliberately create opportunities for the theory to lose. Five prospective programs are already available.

10.1 Electronic Routing Challenge

Use one selected material architecture, one reversible boundary change, one independently compensated carrier condition, and a predeclared ordering among charge, spin, orbital, and dissipative observables. The test fails if the ordering is absent, unstable, or no better than the conventional null after controls.

10.2 Cost-location closure

Choose an intervention that improves one route and instrument every plausible burden channel. The test fails if the predicted displaced cost is absent after uncertainty bounds, hidden reservoirs, and transient storage are addressed.

10.3 Receiver-dependence test

Measure the same physical process with two receiver architectures that TSTOEAO predicts will expose different records. Match sensitivity, bandwidth, timing, and noise. The test fails if the predicted informational difference vanishes.

10.4 Frequency-receiver ledger

For each target, distinguish calculated equivalent frequency, measured resonance, inferred pole, and proposed receiver band. Lock the instrument, bandwidth, expected signature, controls, and failure threshold before acquisition. A missing or control-equivalent signature counts against the specific entry.

10.5 Music and multisensory emotional telemetry

Present musical stimuli in which lyrics and acoustic affect align, conflict, or remain neutral. Compare a lyrics-only model, a population acoustic model, and a personalized longitudinal model using user-confirmed emotion as the reference. The TSTOEAO-derived telemetry claim fails if personalized multisensory calibration does not outperform the word-only baseline across held-out trials.

10.6 Blind record reconstruction

Generate observables from known hidden architectures, conceal the source labels, and compare TSTOEAO-guided reconstruction with standard inverse methods and null models. The record-recoverability extension fails if it adds no reproducible information.

11. Protocol for Future Adversarial Audits

A recurring audit should be published separately from ordinary convergence notes. The following protocol is locked for future use.

31. Define a fixed calendar window before reviewing the papers.
32. Include every candidate encountered in that window, not only the papers ultimately used.
33. Archive the primary source, earliest public preprint, and relevant TSTOEAO proposition.
34. Score source, chronology, mechanism, and discrimination before writing narrative conclusions.
35. Include deliberate null cases and mismatched controls.
36. Use at least one independent scorer who is not invested in promoting TSTOEAO.
37. Publish scorer disagreement and all classification changes.
38. Separate confirmatory analyses from exploratory observations discovered after the scoring lock.
39. Run a multiverse analysis over reasonable scoring thresholds to determine whether conclusions depend upon one arbitrary cutoff.
40. Publish the complete denominator, including rejected, ambiguous, supportive, and contradictory cases.

This protocol follows the central logic of preregistration and Registered Reports: methods and outcome-neutral quality checks should be specified before the result is known. It also adopts the logic of multiverse analysis by exposing how reasonable alternative classification choices affect the conclusion.

12. Consequences for TSTOEAO

The audit improves the theory in five concrete ways.

41. Y is no longer permitted to mean any feature noticed after the fact; a qualifying case must identify a controlled component of relational architecture.
42. Chronology is separated from mechanism, preventing excellent older science from being misrepresented as later confirmation.
43. Source status is separated from conceptual usefulness, preventing reviews and magazine articles from carrying experimental weight they do not possess.
44. Cost-location, receiver dependence, and record recoverability cannot be claimed as demonstrated unless the relevant channels were actually measured.
45. The theory now contains direct failure conditions and a permanent protocol for publishing its own negative ledger.

The audit also clarifies the present evidentiary state. The operational route-selection grammar has strong independent support. The complete ontology of the Encoded Substrate is not uniquely isolated by this audit. The framework’s strongest future evidence will come from tests in which it predicts a result that is not already supplied by local theory, and in which a clear alternative outcome remains possible.

13. Limitations

This is a bounded proponent-run audit, not an independent systematic review. The evidence universe arose from a research process that initially noticed possible convergences, so it is enriched for supportive cases. The scoring rubric was locked within this paper but was not publicly preregistered before the cases were encountered. Several articles were assessed from abstracts and public records rather than complete raw datasets. The score weights are transparent but not externally validated. No direct contradictory experiment was found in the bounded corpus.

These limitations prevent a claim that selection bias has been eliminated. They do not erase the value of the audit. The paper supplies a visible denominator, rejects attractive non-qualifying cases, prevents high mechanistic scores from overriding chronology, and defines the exact prospective procedure required for a stronger test.

14. Conclusion: The Boundary of the Pattern

A theory that sees only confirmations is not ready. A theory that can reject a beautiful match because the chronology is wrong, refuse a magazine article because it is secondary, downgrade a mathematical result because it is not yet empirical, and withhold judgment from a company claim because independent evidence is insufficient has begun to acquire a scientific boundary.

This audit does not show that TSTOEAO has survived every possible attempt at refutation. It shows something more limited and necessary: the framework can distinguish levels of evidence, state automatic rejection conditions, and convert its universal claim into separable propositions that may fail.

The central result is therefore:

TSTOEAO is not strengthened by calling everything confirmation. It is strengthened by identifying exactly what qualifies, what does not, what remains unresolved, and what future observation would force the theory to change.

The previous paper argued that the prediction is the pattern. This paper establishes the other half of the scientific requirement:

A real pattern has a boundary. The theory must be able to say no.

References

1. Swygert, J. Introducing STOEAO – The Swygert Theory of Everything AO. TSTOEAO.com. August 10, 2025. https://tstoeao.com/2025/08/10/introducing-stoeao-the-swygert-theory-of-everything-ao/

2. Swygert, J. Encoded Equilibrium in the Dyadic Manifold: A Unified Framework for Gravity, Magnetism, and Nonlocal Phenomena. TSTOEAO.com. August 10, 2025. https://tstoeao.com/2025/08/10/27/

3. Swygert, J. Chromatic Determinism: Wavelength as Empirical Signature of the Encoded Substrate. TSTOEAO.com. October 28, 2025. https://tstoeao.com/2025/10/28/chromatic-determinism-wavelength-as-empirical-signature-of-the-encoded-substrate/

4. Swygert, J. The Prediction Is the Pattern: Why Repeated Independent Discovery Constitutes Confirmation of a Universal Architecture. July 26, 2026.

5. Popper, K. R. The Logic of Scientific Discovery. London: Hutchinson, 1959.

6. Nosek, B. A., Ebersole, C. R., DeHaven, A. C., and Mellor, D. T. The preregistration revolution. Proceedings of the National Academy of Sciences 115, 2600-2606 (2018). https://doi.org/10.1073/pnas.1708274114

7. Hardwicke, T. E., and Ioannidis, J. P. A. Mapping the universe of registered reports. Nature Human Behaviour 2, 793-796 (2018). https://doi.org/10.1038/s41562-018-0444-y

8. Steegen, S., Tuerlinckx, F., Gelman, A., and Vanpaemel, W. Increasing Transparency Through a Multiverse Analysis. Perspectives on Psychological Science 11, 702-712 (2016). https://doi.org/10.1177/1745691616658637

9. Wang, Y., Zhang, C., Yang, Y., et al. Isomeric multi-hydrogen-bonding enables blue perovskite LEDs. Nature 655, 617-623 (2026). https://doi.org/10.1038/s41586-026-10723-0

10. You, C., Dawkins, R. B., Ferdous, J., et al. Quantum statistical plasmonic metacrystals. Nature 655, 885-891 (2026). https://doi.org/10.1038/s41586-026-10782-3

11. Fujisawa, Y., Wu, P., Nakamura, T., et al. Versatile electronic phases enabled by intertwined multiple frustrations in an antiferromagnetic two-dimensional semimetal. Nature Communications 17, 5981 (2026). https://doi.org/10.1038/s41467-026-75048-y

12. Li, M., Wu, H., Qin, Z., et al. Phonon focusing at room temperature. Nature Physics (2026). https://doi.org/10.1038/s41567-026-03335-y

13. Ruan, X., Ding, C., Cai, H., et al. Bulk polarization fields and interfacial electron sink in MXene-modified iodine-doped Bi4Ti3O12 enhance piezocatalytic H2O2 generation. Nature Communications 17, 3915 (2026). https://doi.org/10.1038/s41467-026-70169-w

14. Zhang, D., et al. Enhanced Lattice Polarization and Directed Charge Transport Toward Pt Surface Sites Accelerate the Volmer Step in Piezocatalytic H2 Evolution on Co-Doped BiFeO3. Advanced Energy Materials (2026). https://doi.org/10.1002/aenm.70895

15. Zhao, M., Zhang, K., Xu, S., et al. Van der Waals surface reconstruction for oriented epitaxial growth of two-dimensional metallic oxides. Nature Materials (2026). https://doi.org/10.1038/s41563-026-02683-7

16. Seiderer, S., Geilen, A., Sliwa, L. N., et al. Giant Brillouin gain in frozen CS2 capillaries. arXiv:2603.25472 (2026). https://doi.org/10.48550/arXiv.2603.25472

17. Calisto, F., Cheung, C., Remmen, G. N., Sciotti, F., and Tarquini, M. On the Inverse Problem in Effective Field Theory. arXiv:2604.15423 (2026). https://doi.org/10.48550/arXiv.2604.15423

18. Yu, M., Bi, H., Lo, H., et al. Nonplanar qubit with tunable gauge symmetry. arXiv:2607.14229 (2026). https://doi.org/10.48550/arXiv.2607.14229

19. Kolodzeiski, P., Gallant, B. M., Richter, L., et al. Alkali-ion-modified zeolitic imidazolate framework glasses. Nature Chemistry (2026). https://doi.org/10.1038/s41557-026-02115-8

20. Qing, Y. M., Shen, Y., Wu, J., Murai, S., Dong, Z., and Okamoto, K. Reconfigurable Giant Nonreciprocity at Near-Normal Incidence via Phase-Change Magneto-Optical Metagratings. Laser & Photonics Reviews (2026). https://doi.org/10.1002/lpor.71438

21. Sun, W., Jiang, Z., Hao, B., et al. Observation of superconductivity-induced leading-edge gap in Sr-doped La3Ni2O7 thin films. arXiv:2507.07409 (2025). https://doi.org/10.48550/arXiv.2507.07409

22. Smith, N. L., Herbert, P. J., Tofanelli, M. A., and Knappenberger, J. A. The Influence of Passivating Ligand Identity on Au25(SR)18 Spin-Polarized Emission. Journal of Physical Chemistry Letters 16, 5168-5172 (2025). https://doi.org/10.1021/acs.jpclett.5c00723

23. CERN. Magnets for the High Luminosity LHC: Small but powerful. 2012. https://home.web.cern.ch/news/news/engineering/magnets-high-luminosity-lhc-small-powerful

24. Lee, S., Lee, S., Jung, S., et al. Broken Kramers Degeneracy in Altermagnetic MnTe. Physical Review Letters 132, 036702 (2024). https://doi.org/10.1103/PhysRevLett.132.036702

25. Foo, J., Mann, R. B., and Zych, M. Relativity and decoherence of spacetime superpositions. arXiv:2302.03259 (2023). https://doi.org/10.48550/arXiv.2302.03259

26. Nature Synthesis. Electrochemical cross-electrophile coupling for carbon-heteroatom and heteroatom-heteroatom bond formation. 2026. https://doi.org/10.1038/s44160-026-01112-6

27. Fedorenko, E., and Varley, R. Language and thought are not the same thing: evidence from neuroimaging and neurological patients. Annals of the New York Academy of Sciences 1369, 132-153 (2016). https://doi.org/10.1111/nyas.13046

28. Damasio, A., and Carvalho, G. B. The nature of feelings: evolutionary and neurobiological origins. Nature Reviews Neuroscience 14, 143-152 (2013). https://doi.org/10.1038/nrn3403

29. Craig, A. D. How do you feel? Interoception: the sense of the physiological condition of the body. Nature Reviews Neuroscience 3, 655-666 (2002). https://doi.org/10.1038/nrn894

30. Weninger, F., Eyben, F., Schuller, B. W., Mortillaro, M., and Scherer, K. R. On the Acoustics of Emotion in Audio: What Speech, Music, and Sound Have in Common. Frontiers in Psychology 4, 292 (2013). https://doi.org/10.3389/fpsyg.2013.00292