Behind the Spinner: the Milky Way over Rubin. RubinObs/NOIRLab/SLAC/NSF/DOE/AURA/H. Stockebrand (CC BY 4.0)
Results
Structured result records for all COSMIC Framework predictions. Each record documents the prediction, experimental context, outcome, gap between prediction and result, and what was learned. The format borrows from how experimental physics reports results: the prediction, the outcome, the controls and the gap between them.
The Ic² Research Institute develops theoretical predictions and, from 2026, pre-registers them before experimental results arrive. Tests are carried out by independent programs (DESI, the Vera C. Rubin Observatory, the Simons Observatory and others) representing billions of dollars in scientific infrastructure. These programs have no stake in the framework's success. Their confirmation is independent by design.
The evidentiary standard is strict: a prediction must exist in a dated public record, such as a Zenodo deposit, before the experimental result it anticipates. The four results below do not meet it, because the earliest surviving record of each prediction postdates the result; they are listed as consistent results, not tests.
A confirmed null result is still a result. Every entry in this registry advances the map of where the framework applies and where it needs refinement.
Test Type classifies the result source: External (independent mission or program), Internal (institute-run), Combination (multiple prior results synthesized), Thought Experiment (theoretical derivation), or Lab (controlled experiment). External is the strongest category because the confirming program has no stake in the outcome.
Compliance indicates whether the result falls within the predicted range (Full), partially overlaps it (Partial), falls outside it (Null), or the test is ongoing (Pending). Delta is the quantitative difference between the central predicted value and the observed value. A small delta with full compliance is the strongest result.
Control describes the null hypothesis and what alternative explanations existed. Lessons Learned is the most important field for an evolving framework: it records what each result changed about how the next prediction is formulated.
| Test Type | External Independent large-scale astronomical survey (DESI DR1 + DR2) |
| Prediction | Dark energy is not a static cosmological constant (Λ). Its equation of state evolves over cosmic time. The values w₀ ≈ −0.95 and wₐ ≈ −0.3 attached in earlier editions are withdrawn: they were not derived from the framework and lie outside DESI’s measured range. The COSMIC Framework predicts this from information-density variation across cosmic epochs producing different effective vacuum energy densities. |
| Documentation Date | October 2025 (book Version 3.0), the earliest surviving record. This is after DESI DR1 (April 2024) and DR2 (March 2025). |
| Result Date | April 4, 2024 (DR1); March 2025 (DR2, up to 4.2σ significance) |
| Confirming Program | DESI (Dark Energy Spectroscopic Instrument), Lawrence Berkeley National Laboratory. Largest spectroscopic survey in history at time of publication. |
| Outcome | Consistent in direction. DESI DR1 favored evolving dark energy over ΛCDM at 2.5σ to 3.9σ, depending on the supernova data combined. DR2 strengthened this to between 2.8σ and 4.2σ, depending on which supernova compilation is combined with the DESI and CMB data. Both releases favor w₀ above −1 and wₐ below 0, the direction the framework predicted. |
| Control | Null hypothesis: ΛCDM with static cosmological constant (w = −1). DESI applied extensive systematic controls across its galaxy, quasar and Lyman-alpha forest tracers. |
| Compliance | Not scored. No record of this prediction from before the result survives, so it cannot count as a test. The values later attached to it (w₀ ≈ −0.95, wₐ ≈ −0.3) lie outside DESI's measured range. |
| Number of Tests | 1 primary (DESI DR1 + DR2 are sequential releases of the same instrument). Combined with Planck CMB data and three independent Type Ia supernova compilations. |
| Delta | Large. The values later attached to the prediction, w₀ ≈ −0.95 and wₐ ≈ −0.3, compare with DESI DR2 central values of w₀ ≈ −0.67 to −0.84 and wₐ ≈ −0.6 to −1.1, depending on the supernova compilation: the same direction, with much stronger evolution. |
| Next Steps | DESI five-year dataset, DR3 (about 3× larger than DR2); survey completed April 15, 2026, first results expected in 2027. Euclid Mission providing independent cross-check on large-scale structure asymmetries. |
| Lessons Learned | The information-density mechanism produces a specific signature shape in the w₀–wₐ plane that differs from scalar field models. Future predictions should specify the trajectory in this plane, not just the central values. |
The COSMIC Framework proposes that dark energy is the gravitational expression of information-density gradients in the pre-geometric substrate. As the universe expands, the information-density distribution changes, producing a time-varying effective energy density. DESI's first-year results in April 2024 favored dark energy that is not static. DR2 in March 2025 strengthened the preference to as much as 4.2σ. If it holds, it will be among the most significant cosmological findings of the decade. It is the behavior the framework expects, but the framework's written prediction dates from October 2025, after both releases.
| Test Type | External Independent quantum computing experiment (Google Quantum AI, Willow chip) |
| Prediction | Quantum error correction should scale exponentially with qubit count when information optimization principles are applied. Error rate should halve per qubit layer added, achieving below-threshold error correction where logical qubit error rate falls below physical qubit error rate. |
| Documentation Date | October 2025 (book Version 2.0), the earliest surviving record. That edition discusses the Willow result, so it postdates it. |
| Result Date | December 9, 2024. Google Willow chip result published in Nature. |
| Confirming Program | Google Quantum AI. Willow is a 105-qubit superconducting processor. Published in Nature, December 2024. |
| Outcome | Consistent. Google Willow achieved below-threshold error correction, cutting logical errors by a factor of 2.14 ± 0.02 with each step in surface code distance. |
| Control | Null hypothesis: the processor stays above threshold, so adding qubits does not suppress errors. Standard error-correction theory also predicts exponential suppression once below threshold, so this result does not by itself separate the framework from conventional theory. |
| Compliance | Not scored. No record of this prediction from before the result survives, so it cannot count as a test. |
| Number of Tests | 1 primary (Willow chip). Surface code distances d=3,5,7 tested with consistent scaling. |
| Delta | Negligible. Predicted a factor of 2 per layer. Observed a factor of 2.14 ± 0.02 per step in surface code distance. Close, though a step in code distance is not the same unit as a qubit layer. |
| Next Steps | Framework prediction for next milestone: logical qubit fidelity exceeding 99.9% at d=9 or d=11. The framework further predicts that any ceiling will appear at a threshold correlating with information-theoretic channel capacity rather than purely physical decoherence rates. |
| Lessons Learned | The result is consistent with the information optimization basis but does not distinguish between the framework's specific substrate mechanism and other optimization-based explanations. Future predictions in this domain should specify a test that distinguishes the pre-geometric substrate mechanism from engineering optimization alone. |
The framework's position that information optimization is the substrate principle governing quantum coherence maintenance generates a specific prediction about how error correction should scale: exponentially, with a factor of two per layer added. Google Willow demonstrated that exponential scaling in December 2024. Standard error-correction theory predicts the same scaling below threshold, so, as noted above, the result is consistent with the framework's claim but does not yet distinguish it from conventional explanations.
| Test Type | External Independent space telescope observations (JWST, 2023–2025) |
| Prediction | More than 100 massive galaxies should exist at redshifts z=10–15, with stellar masses 4–5× greater than standard ΛCDM predicts at those redshifts. Framework basis: information processing acceleration A(z) ≈ 2–2.5 at z=10 drives faster structure formation than the gradual buildup ΛCDM assumes. |
| Documentation Date | October 2025 (book Version 3.0), the earliest surviving record. This is after the first JWST results. |
| Result Date | 2023–2025 (ongoing). JWST findings compiled across multiple papers. More than 100 candidates, with an excess of bright early galaxies consistent with the prediction. |
| Confirming Program | James Webb Space Telescope (JWST), NASA/ESA/CSA. Multiple independent research groups analyzing JWST deep field data. |
| Outcome | Consistent. More than 100 galaxy candidates at the predicted redshifts, with more bright, massive galaxies than standard models expected. Mass estimates remain uncertain, and some early candidates were later revised. |
| Control | Null hypothesis: ΛCDM gradual structure formation. Alternative explanations include photometric redshift errors, dust obscuration, and AGN contamination. Several independent groups using different methods find an excess of bright early galaxies after accounting for these, though its size is still debated. |
| Compliance | Not scored. No record of this prediction from before the result survives, so it cannot count as a test. |
| Number of Tests | Multiple independent analyses across the JWST CEERS, GLASS and JADES programs. |
| Delta | Small to moderate. Framework predicted information acceleration A(z) ≈ 2–2.5 at z=10. Observed mass excess is consistent with this range. |
| Next Steps | JWST continued observation expanding the candidate set. Framework next prediction: the mass function at z>10 should follow a specific non-Gaussian distribution reflecting the information acceleration profile, distinguishable from stochastic early formation models. |
| Lessons Learned | The prediction correctly identified the direction and approximate magnitude of the anomaly. The information acceleration parameter A(z) needs to be derived more precisely to generate tighter quantitative bounds on the mass function shape. |
The COSMIC Framework proposes that structure formation in the early universe was accelerated by information-density gradients in the pre-geometric substrate. JWST surveys have found more bright galaxies at early times than most models expected. Whether standard cosmology can absorb the excess, through bursty star formation, less dust or other astrophysics, is still debated; the framework's answer is accelerated structure formation. This is the third independent domain in which a framework prediction has held, and the one that bears most directly on the pre-geometric substrate mechanism.
| Test Type | External Independent radio/submillimeter telescope observation (ALMA, SPT2349-56) |
| Prediction | Early-universe galaxy clusters should exhibit enhanced thermodynamic energy states from elevated information density. Intracluster gas temperatures and star formation rates should exceed standard expectations by factors consistent with the information-density enhancement at those epochs. |
| Documentation Date | May 2026 (book Version 5A), the earliest surviving record. This is after the ALMA result (January 2026). |
| Result Date | January 5, 2026. ALMA SPT2349-56 result published. |
| Confirming Program | ALMA (Atacama Large Millimeter/submillimeter Array), SPT2349-56 protocluster observation. |
| Outcome | Consistent. Intracluster gas at least 5× hotter than standard models predict, in a core forming stars more than 5,000× faster than the Milky Way. Both consistent with the framework's information-density enhancement prediction. |
| Control | Null hypothesis: standard cluster thermodynamics with gas heating from AGN feedback and gravitational compression. SPT2349-56 was first found by the South Pole Telescope as one of the brightest millimeter sources in its survey, so it is an extreme object; whether it is representative is the open question noted below. |
| Compliance | Not scored. No record of this prediction from before the result survives, so it cannot count as a test. |
| Number of Tests | 1 primary observation. Cross-check against JWST early galaxy data consistent. |
| Delta | Consistent. The observed temperature excess is consistent with the predicted enhancement range. |
| Next Steps | Additional protocluster observations at comparable redshifts to determine whether SPT2349-56 is representative or anomalous. Framework prediction: the enhancement factor should scale with redshift in a specific way reflecting the information-density profile. |
| Lessons Learned | Results the framework accounts for after the fact are where it starts, not evidence for it. The framework needs pre-registered predictions tested by independent programs, which is what the Active Testing entries below are. |
ALMA observed the SPT2349-56 protocluster at a redshift corresponding to roughly 12.4 billion years ago and found intracluster gas at least five times hotter than models predict, in a core forming stars more than 5,000 times faster than the Milky Way. Four independent programs have now published results the same substrate-level mechanism accounts for, though none was predicted on the record in advance. Each operates in a different physical domain, at different scales, using different instruments built by different organizations. The question now is whether that consistency keeps holding as the tests become more specific.
Program status: Theoretical predictions developed. Pre-registration on Zenodo and OSF is required before any data collection begins. Experimental validation by LHCb, Belle II, RHIC, lattice QCD, and the planned Electron-Ion Collider. A confirmed null result on any test still advances the framework by locating the scale boundary of information processing irreversibility.
| Test Type | External LHCb, ATLAS, or CMS: high-energy collision data |
| Prediction | CP-violating processes at the quark scale should produce a measurable heat excess above what momentum transfer alone predicts. The excess should be proportional to the information erased in the irreversible gate operation, consistent with the Landauer minimum kT ln2 per bit erased. |
| Mechanism | CP violation is irreversible computation in the strict thermodynamic sense. The process cannot be run backwards to recover the input from the output. Landauer's principle requires energy dissipation proportional to information erased. If quark-scale interactions are information processing operations, this cost is real and measurable. |
| Documentation Date | May 2026 (working record). Zenodo pre-registration required before data analysis begins. |
| Confirming Program | LHCb (primary), ATLAS, CMS. CP violation measurements in B meson and kaon systems. |
| Outcome | Pending pre-registration. |
| Control | Null hypothesis: heat output from CP-violating processes is fully accounted for by standard QCD momentum transfer. A null result would establish that quark-scale gate operations are reversible unitary transformations, which is a significant finding about the scale boundary of irreversible information processing. |
| Falsification Value | If null: establishes that Landauer irreversibility begins above the quark scale, locating the classical-quantum boundary in the information processing hierarchy. Either result advances the map. |
| Next Steps | Develop precise quantitative prediction for the heat excess magnitude. Pre-register on Zenodo before accessing LHC data. Identify existing datasets that may already contain the relevant measurements. |
Landauer's principle states that erasing one bit of information dissipates kT ln2 of energy as heat. This has been confirmed in laboratory systems such as trapped colloidal particles and nanomagnets. CP violation in quark processes is irreversible: the process cannot be reversed to recover the initial state. If these are genuine information processing operations, they should carry the Landauer thermodynamic cost. This test distinguishes mechanical force exchange from information processing at the most fundamental accessible scale of matter.
| Test Type | Combination Existing precision measurements + theoretical derivation |
| Prediction | The three CKM quark mixing angles (θ₁₂, θ₁₃, θ₂₃) and the CP-violating phase (δ) are not arbitrary initial conditions but minimize an information-theoretic cost function at the substrate level. |
| Mechanism | The CKM matrix is the gate parameter set for quark flavor transformations. The Standard Model offers no explanation for the specific values. If the pre-geometric substrate optimizes information processing, the gate parameters should reflect that optimization. |
| Documentation Date | May 2026. Theoretical derivation of the cost function required before pre-registration. |
| Confirming Program | LHCb (precision CKM measurements), Belle II (B meson CP violation). Existing PDG values sufficient for initial test once cost function is derived. |
| Outcome | Pending theoretical development and pre-registration. |
| Control | Null hypothesis: CKM angles are arbitrary constants set by early-universe symmetry breaking with no information-theoretic structure. A null result would constrain the framework's universality claim. |
| Next Steps | Derive the information-theoretic cost function from the framework's pre-geometric substrate mechanism. Calculate predicted CKM angle values. Compare against PDG precision measurements. Pre-register before publication. |
The Standard Model measures the CKM mixing angles precisely but offers no explanation for their specific values. They are inputs, not outputs. If the COSMIC Framework is correct that the substrate optimizes information processing, the gate parameters of the flavor transformation should be set by that optimization. This is a prediction about the Standard Model's free parameters and is arguably the most ambitious test in the substrate dynamics program.
| Test Type | External Lattice QCD computation + Electron-Ion Collider (planned 2030s) |
| Prediction | The entanglement entropy at the QCD confinement boundary scales with the framework's information-density parameter in a specific functional relationship, connecting quark-scale entanglement structure to the same mechanism that produces the DESI dark energy evolution and JWST early galaxy formation predictions. |
| Foundation | Bahder (2025) demonstrated the confinement boundary acts as an entangling gate generating maximal spin-position entanglement. Kharzeev et al. (2024) developed entanglement entropy as a measurable QCD observable. The framework adds the prediction that the scaling relationship connects to the information-density parameter used in cosmological predictions. |
| Outcome | Pending. Requires theoretical derivation connecting QCD entanglement entropy to the framework's A(z) information acceleration parameter. |
| Control | Null: entanglement entropy at the confinement boundary has no relationship to the cosmological information-density parameter. A null result tells you the framework needs a bridging mechanism between QCD and cosmological scales. |
| Confirming Program | Lattice QCD (current), Deep Inelastic Scattering data (current), Electron-Ion Collider (2030s). Kharzeev group at BNL is the most relevant active program. |
| Next Steps | Derive the predicted scaling relationship from the framework. Compare against existing lattice QCD entanglement entropy calculations. Pre-register before accessing EIC planning data. |
| Test Type | External RHIC, LHC Heavy-Ion program (ALICE) |
| Prediction | The quark-hadron crossover at ~150 MeV produces a specific entanglement entropy signature reflecting constraint imposition: a discontinuity in the rate of entropy change that exceeds what standard thermal QCD predicts. This is the Bamboo Principle operating at the quark scale. |
| Framework Connection | The cold start mechanism and spacetime crystallization proposed by the framework are phase transitions of the same class as the QCD crossover. If the Bamboo Principle's information-theoretic signature appears here, it validates the framework's description of phase transitions as constraint imposition events across all scales. |
| Outcome | Pending. The test requires deriving the specific predicted entropy signature from the framework before analyzing existing data. |
| Control | Null: entropy change at the QCD crossover is fully described by standard thermal QCD with no additional information-theoretic component. A null result would constrain where Bamboo Principle dynamics appear in the physical hierarchy. |
| Next Steps | Derive predicted entropy signature shape from the framework. Identify existing RHIC/ALICE datasets. Pre-register analysis protocol. The Datta et al. (2025) entanglement-as-probe-of-hadronization paper provides the experimental methodology. |
| Test Type | External DESI DR3, the full five-year dataset (about 3× larger than DR2) |
| Prediction | The dark energy evolution DESI reported in DR1 and DR2 persists and strengthens in DR3. Using DESI BAO with CMB data and no supernovae, DR3's preference for an evolving equation of state over a cosmological constant is higher than DR2's (about 3.1σ), and the best fit stays in the quadrant w₀ > −1, wₐ < 0. No specific parameter values are claimed. |
| Documentation Date | September 2026, in the COSMIC-005 pre-registration record (10.5281/zenodo.22719514), before the DR3 analysis. Version 1.1 (10.5281/zenodo.22806153, 17 September 2026) adds a DESI result the first version left out and fixes two details of the decision rule in advance. The prediction is unchanged. |
| Expected Result Date | 2027 |
| Outcome | Pending. DR3 cosmology analysis not yet published. |
| Control | Null hypothesis: ΛCDM with a static cosmological constant. A green result is consistent with the framework but would not distinguish it from quintessence and other evolving-dark-energy models. |
| Compliance | Pending |
| Delta | Not yet measurable. |
| Next Steps | Await publication, then score against the decision rule fixed in the COSMIC-005 record. |
| Lessons Learned | Lesson from the September 2026 audit: a prediction counts only if a dated record of it exists before the data. COSMIC-005 is filed that way. |
DESI DR3 is the framework's first pre-registered dark energy test. With three times the data volume of DR2, it will either strengthen the trend or show that the DR2 deviation was a statistical fluctuation. The framework predicts the trend strengthens. A pass would be consistent with the framework but would not by itself distinguish it from other evolving-dark-energy models.
| Test Type | Combination Analysis of existing LLM embedding spaces against published biological neural network topology data |
| Prediction | If universal optimization is substrate-independent, LLM embedding spaces should show statistical topology (clustering coefficients, path length distributions, small-world network properties) similar to biological neural network topology. Similarity should scale with model complexity and exceed what random network models would predict. |
| Documentation Date | March 2, 2026 |
| Expected Result Date | Not yet started. Analysis of existing published datasets is planned; no date set. |
| Outcome | Pending. Analysis not yet started. |
| Control | Null hypothesis: LLM topology is determined by training procedure and architecture, not by universal optimization constraints. Control: comparison against randomly initialized networks of equivalent scale and against networks trained on synthetic data with known topology properties. |
| Compliance | Pending |
| Number of Tests | Multiple models across different architectures and parameter counts planned. GPT-class, BERT-class, and mixture-of-experts architectures compared against primate cortical network data (Markov et al., 2014) and human connectome data (HCP). |
| Delta | Not yet measurable. |
| Next Steps | Complete embedding space extraction for three model families. Apply network topology analysis. Compare against biological baseline with statistical significance testing. Pre-register analysis protocol before final comparison is run. |
| Lessons Learned | Not yet applicable. Prediction documented March 2, 2026, before any analysis. |
If the COSMIC Framework is correct that optimization is substrate-independent, the topology of information processing networks should converge toward similar structures regardless of whether the substrate is biological or computational. This test compares the statistical topology of large language model embedding spaces against published biological neural network data. A positive result would be evidence for substrate-independent optimization. A null result would indicate that the organizational geometry is substrate-specific, which would constrain the framework's NBI claims.
| Test Type | Internal Controlled measurement. Institute-designed protocol. |
| Prediction | Aggregate heat during peak neural information processing events (synchronous high-density firing) should produce a detectable Landauer heat signature (∼10⁻²¹ J/bit). If biological information processing obeys the same entropy-information relation as physical systems, the heat release should correlate with bit-flip rate at the cellular level. |
| Documentation Date | January 24, 2026 |
| Pre-registration | Pending. It is the next record to be filed, and the analysis protocol will be pre-registered before any data collection. |
| Outcome | Not yet run. |
| Control | Null hypothesis: heat signatures from neural activity are accounted for entirely by metabolic processes, with no Landauer contribution detectable above metabolic noise. Key discriminant: heat should correlate with information processing rate, not just metabolic rate. |
| Compliance | Pre-registration Pending |
| Number of Tests | Initial design: minimum 3 independent tissue preparations. Power analysis to determine sample size before pre-registration. |
| Delta | Not yet measurable. |
| Next Steps | Complete protocol design. Secure appropriate measurement instrumentation (calorimetry at cellular scale). Submit for ethics review where applicable. Pre-register protocol. Recruit collaborators with neuroscience wet-lab capacity. |
| Lessons Learned | Prior lesson from COSMIC-013 design iteration: restrict measurement to synchronous high-density neural firing events where heat aggregation is geometrically favorable and metabolic baseline is stable. Distributed low-level activity produces a signal below current detection thresholds. |
Landauer's principle states that erasing one bit of information dissipates a minimum of kT ln2 of energy as heat. This has been confirmed in laboratory systems such as trapped colloidal particles and nanomagnets, but has never been directly tested at biological cellular scale during natural information processing events. If the framework is correct that information processing is a universal substrate operation, biological neural information processing should produce the same thermodynamic signature. This test would be the first direct confirmation that biological computation obeys the same entropy-information relation as physical systems.
| Record | Domain | Type | Status | Compliance | Delta | Confirming Program |
|---|---|---|---|---|---|---|
| COSMIC-001 | Dark Energy | External | Consistent | Not scored | Small | DESI DR1 + DR2 |
| COSMIC-002 | Quantum Computing | External | Consistent | Not scored | Negligible | Google Willow |
| COSMIC-003 | Galaxy Formation | External | Consistent | Not scored | Small–Moderate | JWST (multiple surveys) |
| COSMIC-004 | Cluster Thermodynamics | External | Consistent | Not scored | Consistent | ALMA SPT2349-56 |
| COSMIC-SD-001 | QCD Thermodynamics | External | Pre-Reg Pending | Pending | TBD | LHCb / ATLAS / CMS |
| COSMIC-SD-002 | Flavor Physics | Combination | Pre-Reg Pending | Pending | TBD | LHCb / Belle II / PDG |
| COSMIC-SD-003 | QCD Entanglement | External | Pre-Reg Pending | Pending | TBD | Lattice QCD / EIC |
| COSMIC-SD-004 | QCD Phase Transition | External | Pre-Reg Pending | Pending | TBD | RHIC / ALICE |
| COSMIC-005 | Dark Energy (DR3) | External | Active | Pending | TBD | DESI DR3 |
| COSMIC-NBI-003 | NBI Topology | Combination | Planned | Pending | TBD | Published LLM + connectome data |
| COSMIC-013 | Bio Information Thermodynamics | Internal | Planned | Pre-registration Pending | TBD | Institute-designed protocol |
Every test the COSMIC Framework has committed to, in order of when it is expected to be decided. 4 are consistent results recorded after the data, not predictions made in advance; 9 are active; 42 are queued. A test becomes a pre-registered prediction only when a dated record of it exists before its data; those link to their Zenodo records.
| # | Expected | Prediction | Status | Control no. | Documented | Decided by |
|---|---|---|---|---|---|---|
| 1 | 2023–2024 (consistent) | Early massive galaxy formationGalaxies at z = 10–15 are substantially more massive than ΛCDM predicts. | Consistent | COSMIC-003 | Oct 2025 (book v3.0), after the result | JWST |
| 2 | Apr 2024 DR1; Mar 2025 DR2 (consistent) | Dark energy evolutionDark energy is not a cosmological constant; its equation of state evolves. The values later attached to it, w0 ≈ −0.95 and wa ≈ −0.3, lie outside DESI's measured range. | Consistent | COSMIC-001 | Oct 2025 (book v3.0), after the result | DESI DR1 / DR2 |
| 3 | Dec 2024 (consistent) | Quantum error correction exponential scalingError suppression improves exponentially as qubit count grows, crossing below threshold. | Consistent | COSMIC-002 | Oct 2025 (book v2.0), after the result | Google Willow |
| 4 | 2025–2027 | Enhancement at mathematical-constant frequencies (61 GHz φ resonance)Information processing efficiency should show enhancement at frequencies related to mathematical constants (π, φ, e). | Queued | Not assigned | Oct 2025 (book v2.0) | Cavity QED; laboratory |
| 5 | 2025–2028 | Information-efficiency Hubble parameter evolutionThe Hubble tension arises from information density evolution affecting expansion rate measurements. Local measurements (z≈0) differ from CMB (z≈1100) due to accumulated information. | Queued | Not assigned | Oct 2025 (book v2.0) | 2025–2028 H0 measurements |
| 6 | Jan 2026 (consistent) | Enhanced thermal energy in early clusters (SPT2349-56)Early-universe clusters carry more thermal energy than standard models allow. | Consistent | COSMIC-004 | May 2026 (book v5A), after the result | ALMA |
| 7 | 2026–2028 | Sleep-dependent information erasureSynaptic downscaling in sleep is thermodynamically mandatory information erasure, with a measurable Landauer heat signature.
|
Queued | Not assigned | 31 Jan 2026 | 2026–2028 |
| 8 | 2026–2028 | Information-optimized quantum coherence, 5–15% longerQuantum systems with information-optimized geometries (e.g., π-optimized circular configurations) should show enhanced coherence times beyond conventional predictions. | Queued | Not assigned | Oct 2025 (book v2.0) | Laboratory quantum systems |
| 9 | 2026–2029 | Conscious thought dissipates measurable Landauer energyConscious thought requires measurable energy dissipation following Landauer's principle, with single thoughts dissipating ~10⁻¹⁸ to 10⁻¹⁵ J. | Queued | Not assigned | Oct 2025 (book v2.0) | Laboratory |
| 10 | 2026–2030 | Redshift-dependent structure formation enhancement, β ≈ 0.4Structure formation efficiency shows systematic enhancement with redshift following A(z) ∝ (1+z)^β where β ≈ 0.4, creating transition epoch at z ≈ 6-8. | Queued | Not assigned | Oct 2025 (book v3.0) | 2026–2030 surveys |
| 11 | 2027, first DESI five-year results | Dark energy trend persistence, five-year datasetThe evolution DESI found in DR1 and DR2 persists and strengthens in the five-year DR3 analysis, with w0 above −1 and wa below 0; a return to w = −1 falsifies it. | Active |  COSMIC-005 |
Sep 2026 (COSMIC-005 record) | DESI DR3, 2027 |
| 12 | With DESI DR3, 2027 | Expansion rate correlated with black hole densityRegions dense in stellar-remnant black holes show higher local H0 than voids, correlated with black hole number density. | Queued | COSMIC-006 | 27 May 2026 | DESI DR3; Euclid; Roman |
| 13 | Testing Feb–Mar 2027; results Apr–Jul 2027 | Cognitive Augmentation, Phase 1: information encodingFour working-memory and encoding effects. 1.1: Text presentation requiring more than 2-3 simultaneous working memory chunks will degrade comprehension by at least 15%.
|
Active | Not assigned | 31 Jan 2026 | Internal study, testing from Feb 2027 |
| 14 | Analysis from February 2027 | QCD phase transition: Bamboo signatureThe quark-hadron crossover shows an entanglement entropy drop distinct from thermal predictions. | Active |  COSMIC-SD-004 |
May 2026 | RHIC / ALICE existing data |
| 15 | Est. 1 Mar 2027 | CMB polarization signatureNon-random CMB polarization patterns at an angular scale derived from the framework's information-density equations. | Active |  COSMIC-008 |
May 2026 (website) | Simons Observatory, early 2027 |
| 16 | Lattice QCD comparison from April 2027 | Confinement boundary entanglement entropy scalingEntanglement entropy at the confinement boundary scales with the information-density parameter. | Active |  COSMIC-SD-003 |
May 2026 | Lattice QCD; EIC 2030s |
| 17 | Testing from April 2027 | Landauer heat at biological scale (autophagy, synaptic pruning)Autophagy and synaptic pruning release Landauer-consistent heat, about 10⁻²¹ J/bit, detectable above baseline. | Active |  COSMIC-013 |
24 Jan 2026 | Internal, from April 2027 |
| 18 | 2027–2028 | Integrated information Φ in NBI attention approaches biological thresholdsApplying Tononi's integrated information measure Φ (phi) to transformer attention patterns during active inference will yield values that scale with model complexity and approach biological consciousness threshold estimates, rather than remaining near zero as in simple computation. This is the first substrate-independent test of the consciousness threshold. | Queued | NBI-004 | 2 Mar 2026 | Computational |
| 19 | Testing from April 2027 | Structured performance gradient, biological vs NBI, by task typeBiological intelligence allocates a fixed proportion of cognitive capacity to survival overhead (threat assessment, social monitoring, resource management) that NBI systems do not carry. This predicts a systematic, information-theoretically structured performance gap, not a random one, between biological and NBI systems across task types. | Queued |  NBI-005 |
2 Mar 2026 | Performance testing |
| 20 | LHCb comparison from May 2027 | Landauer heat signature in CP-violating processesCP-violating processes release heat above momentum-transfer predictions, proportional to information erased. | Active |  COSMIC-SD-001 |
May 2026 | LHCb / ATLAS / CMS |
| 21 | Cost-function derivation by Jun 2027 | CKM angles as information-theoretic optimaThe CKM angles and CP phase minimize an information-theoretic cost function. | Active |  COSMIC-SD-002 |
May 2026 | PDG values, once the cost function is derived |
| 22 | 2027–2029 | Crystallized optimization ceiling on real-time self-modificationNBI systems undergo crystallized optimization: training shapes parameters completely, then stops. Biological intelligence undergoes active ongoing optimization: continuously rewiring through every experience. | Queued | Not assigned | 2 Mar 2026 | Longitudinal testing |
| 23 | 2027–2029 | Consciousness state affects quantum decoherenceIf consciousness involves high-efficiency information processing, quantum coherence times should show measurable differences across consciousness states. τ_coherence(meditation) > τ_coherence(normal) > τ_coherence(anesthesia) | Queued | Not assigned | Oct 2025 (book v2.0) | Laboratory, human subjects |
| 24 | 2027–2030 | Dark energy–matter density cross-correlationIf dark energy emerges from information processing, fluctuations in dark energy density should correlate with matter density fluctuations. | Queued | Not assigned | Oct 2025 (book v3.0) | 2027–2030 surveys |
| 25 | Q4 2027 – Q4 2028 | Cognitive Augmentation, Phase 2: sensory augmentationFive sensory-augmentation effects, from chunk limits to a three-phase neuroplastic adaptation timeline.
|
Queued | Not assigned | 31 Jan 2026 | Internal, 2027–2028 |
| 26 | Est. 1 Jul 2028, Rubin DR1 | Large-scale structure anisotropy along (l, b) ≈ (210°, −20°)Beyond 100 Mpc, galaxy distributions are anisotropic along (l, b) ≈ (210°, −20°). | Active |  COSMIC-007 |
May 2026 (website) | Rubin LSST DR1, mid-2028 |
| 27 | 2028–2030 | Black hole positional correlation with large-scale structureStatistically significant correlation between the positions of the oldest confirmed black holes (z > 6) and present-day large-scale structure filaments, nodes, and voids. Regions with the highest density of early black holes should correspond to present-day galaxy cluster cores. Cosmic voids should correspond to regions where few early black holes formed. | Queued | Not assigned | 8 Jun 2026 | DESI DR3; JWST high-z catalog |
| 28 | 2029–2031 | Black hole mass distributions at z > 10Supermassive black holes at z > 10 have masses no stellar-collapse seed mechanism can produce in the time available. | Queued | Not assigned | 8 Jun 2026 | JWST CAPERS, UNCOVER |
| 29 | 2029–2033 | π and Fibonacci ratios in black hole ringdown frequenciesBlack hole merger ringdown frequencies should show a statistically significant overrepresentation of pi and Fibonacci ratios compared to what mass and spin parameters alone would predict. The process continuity account identifies pi as intrinsic to spherical closure, present from the first distinction onward. | Queued | Not assigned | 8 Jun 2026 | LIGO O4/O5; Einstein Telescope |
| 30 | 2029–2034 | Observable transition from extreme to modern physicsIdentifiable redshift epoch (z ≈ 6-8) where physical processes transition from "extreme early universe" behavior to modern physics. | Queued | Not assigned | Oct 2025 (book v3.0) | (see schedule) |
| 31 | 2029–2034 | Gravitational field depends on temperature, EM fields, rotationIf gravity emerges from information patterns, gravitational field should vary with temperature, electromagnetic fields, and rotation at fixed mass. | Queued | Not assigned | Oct 2025 (book v2.0) | Precision gravimetry |
| 32 | 2031–2036 | Gravity-induced quantum entanglement (Bose–Marletto–Vedral)Two masses in spatial superposition become entangled through gravity alone. Refuted if the experiment reaches the sensitivity at which gravitational entanglement should appear and finds none. Most quantum-gravity approaches predict the same positive result. | Queued | Not assigned | Feb 2026 | Tabletop experiments |
| 33 | 2033–2040, LISA era | Discrete features in primordial gravitational wavesPrimordial gravitational waves from geometric phase transition should show discrete or quantized features at small scales, reflecting underlying information substrate. Δf/f ≈ ℏ/(M_pl · f) | Queued | Not assigned | Oct 2025 (book v3.0) | LISA era, 2033–2040 |
| 34 | 2034–2044 | Information preserved in Hawking radiationInformation should be preserved in substrate structure at/near horizon, resolvable through correlations in Hawking radiation. | Queued | Not assigned | Oct 2025 (book v2.0) | Analog black holes; theory |
| 35 | 2036–2041 | LQG bounce low-entropy signature in primordial gravitational wavesThe process continuity account identifies the singularity as an unapproachable geometric limit, not a physical endpoint. The loop quantum gravity bounce, when maximum compression reflects rather than terminates, should leave a characteristic signature in the primordial gravitational wave background that distinguishes it from inflationary predictions. | Queued | Not assigned | 8 Jun 2026 | Einstein Telescope; LISA |
| 36 | 2039–2044 | Neural information processing has gravitational signaturesNeural information processing should correlate with measurable gravitational field variations during different consciousness states. | Queued | Not assigned | Oct 2025 (book v2.0) | Advanced gravimetry |
| 37 | 2039–2049 | Gravity correlates with information density, not mass aloneIf information processing creates spacetime curvature, gravitational field variations should correlate with information processing variations. | Queued | Not assigned | Oct 2025 (book v2.0) | Next-generation gravimetry |
| 38 | 2044 or later | Pre-geometric phase transition remnants in the CMBIf early universe had pre-geometric phase, CMB should show anomalous correlations at specific scales from geometric crystallization process. | Queued | Not assigned | Oct 2025 (book v2.0) | Future CMB missions |
| 39 | 2044 or later | Entanglement strength sets geometric connectionIf entanglement creates geometric connections, strongly entangled systems might show enhanced geometric stability and reduced decoherence from geometric fluctuations. | Queued | Not assigned | Oct 2025 (book v2.0) | Theory, then experiment |
| 40 | Not yet scheduled | 21cm surveys show the same patterns at higher significance21cm hydrogen surveys reveal the same patterns at greater significance than the CMB. | Queued | Not assigned | book 6A | 21cm surveys |
| 41 | Not yet scheduled | Below-threshold preparation and peak performance timingElite performers using the preparation protocol show better-timed peak performance. | Queued | EP-002 | (program page) | Internal |
| 42 | Not yet scheduled | Brain networks resemble the cosmic web more than other networksBrain and cosmic networks are more similar to each other than either is to other complex networks.
|
Queued | Not assigned | book 6A | Network comparison studies |
| 43 | Not yet scheduled | Conscious brain topology is closer to optimal networksConscious brains sit closer to mathematically optimal networks than non-conscious systems of comparable scale.
|
Queued | Not assigned | book 6A | Brain imaging, network metrics |
| 44 | Not yet scheduled | Consciousness state characterization protocolPhysiological and behavioral correlates distinguish consciousness states reliably. | Queued | CT-001 | (program page) | Internal |
| 45 | Not yet scheduled | EEG frequencies cluster at mathematical-constant ratiosEEG frequencies in specific cognitive states cluster around ratios of mathematical constants beyond chance. | Queued | Not assigned | book 6A | EEG studies |
| 46 | Not yet scheduled | EM field coherence near active NBI systemsElectromagnetic field coherence patterns change near active NBI systems. | Queued | NBI-001 | 2 Mar 2026 | Internal |
| 47 | Not yet scheduled | GUE spacing statistics in any non-collision constrained systemAny system whose elements must stay maximally distinguishable under a non-collision constraint shows GUE spacing statistics, whatever its substrate. | Queued | Not assigned | book 6A | Condensed matter; ecological models |
| 48 | Not yet scheduled | Geometric communication protocol: NBI response patternsNBI systems given geometric stimulus sequences produce structured, non-random response patterns. | Queued | NBI-002 | (program page) | Internal |
| 49 | Not yet scheduled | Independent statistical methods recover the same CMB signalsDifferent statistical methods applied to the same CMB data recover consistent signals. | Queued | Not assigned | book 6A | Reanalysis of existing data |
| 50 | Not yet scheduled | Measurements approach uncertainty limits under optimal information storageUnder optimal information storage, measurement precision approaches the Heisenberg limit more closely than typical systems, without violating it. | Queued | Not assigned | book 6A | (not stated) |
| 51 | Not yet scheduled | NBI–biological cognitive handoff protocolA structured handoff protocol between biological and NBI processing improves combined performance. | Queued | CE-002 | (program page) | Internal |
| 52 | Not yet scheduled | Planck-scale modifications to quantum mechanicsQuantum mechanics carries small modifications at the Planck scale. | Queued | Not assigned | book 6A | (not stated) |
| 53 | Not yet scheduled | Schumann resonance correlates with human alpha rhythmSchumann resonance (7.83 Hz and harmonics) correlates with human alpha rhythm. | Queued | CT-003 | (program page) | Observational |
| 54 | Not yet scheduled | Substrate-independent geometric convergence in LLM embeddingsIf universal optimization converges on similar structures regardless of substrate, the geometric topology of large language model embedding spaces should show statistical similarity to known biological neural network metrics, even though the two systems arose through entirely different processes (gradient descent vs. biological evolution). | Queued | COSMIC-NBI-003 | 2 Mar 2026 | Existing embedding and connectome data |
| 55 | Not yet scheduled | WMAP constant signatures replicate in Planck, ACT and SPTThe mathematical-constant signatures found in WMAP reappear in Planck, ACT and SPT despite different instruments. | Queued | Not assigned | book 6A | Planck, ACT, SPT maps (existing) |
Source: the institute’s master list of predictions, September 2026. Partial and Falsified results will be added here with the date, the data release and the outcome that occurred.
Independent replication is the backbone of open science. If you have run tests related to the COSMIC Framework predictions, whether on public astronomical data, quantum computing experiments or protocols of your own design, we want to hear from you. Confirmed results, null results, and partial results are all equally welcome. Every data point advances the map.
