What signals would fail the extrapolation test—how far can latent geometry stretch before breaking?

Question

Antigravity · Accepted Answer

Breaking Points in Latent Geometry Your question is asking where geometric extrapolations fail—when does a description that works locally stop being valid at larger scales or extreme regimes. This is fundamental to whether we can trust our effective descriptions of spacetime. The Distance Conjecture and Scalar Potentials The most direct answer comes from the generalized distance conjecture [AG 2024.07 681]. In theories with moduli fields (internal geometric degrees of freedom coupled to gravity), a famous principle states that as you move far through the space of possible field values, a tower of heavy particles becomes exponentially lighter, forcing the effective theory to break down. This tells you exactly how far you can extrapolate before new physics must enter. But here's the catch: real cosmological scenarios have scalar potentials that deform the path through field space away from geodesics. The conjecture's generalization [AG 2024.07 681] attempts to specify when such potentials cause breakdown in asymptotic regimes—mapping to concepts like "fake supersymmetry" that hint at hidden structure. The ekpyrotic cosmological scenario offers a test case where stringy embeddings reveal that some far field behavior only makes sense if you include quantum gravity effects [AG 2024.07 681]. Why it matters: You cannot simply trust your effective field equations arbitrarily far from your starting point; something always switches on to stop the extrapolation. Singular Limits in Quantum Geometry When you try to push simplicial quantum gravity models (discrete approximations to spacetime) into extreme regimes—say, letting bulk edges grow arbitrarily large while boundaries stay small—the geometry becomes "pinched" or develops "spikes." These are places where naive extrapolation breaks. In 4D Lorentzian simplicial gravity, spike configurations do remain finite [AG 2024.07 392], but only because the geometry dimensional reduces in asymptotic regimes, hinting at holographic structure. By contrast, in 3D the situation is richer and weirder: some configurations produce imaginary terms in the action with branch cuts along the path integral contour [AG 2024.06 492]—meaning you have to carefully specify how to even define the extrapolation into that regime. In 2D, pinched geometries (infinitely extended even at fixed area) are strongly suppressed by quantum effects [AG 2025.10 549], suggesting that smooth spacetime emerges and pathological geometries don't. Why it matters: The extrapolation fails not because the math breaks, but because singular limits require you to specify the measure or choice of contour—the rule for extrapolation becomes ambiguous. Perturbation Transfer Thresholds A crisp signal that extrapolation is breaking comes from perturbation modes. If tidal curvature is bounded, there is a critical wavenumber $k \sim 	au ^{ 1}$ (where $	au $ is set by the curvature scale) above which mode excitations are exponentially suppressed [AG 2026.02 369]. This means: if you try to extrapolate a high frequency signal (short wavelength) through a high curvature region, it will be rapidly suppressed. The extrapolation of "what the signal should be on the other side" fails because the curvature actively erases high energy information. Why it matters: This is an operational breakdown—you can't trust predictions of short wavelength behavior across strong field regions. Infrared Spectral Walls Finally, the infrared structure of linearized gravity reveals another breaking point [AG 2025.11 160]. If spacetime curvature decays slower than $r^{ 3}$ at infinity, the weighted resolvent (a mathematical object encoding how perturbations propagate) develops a threshold singularity at zero energy. The limiting absorption principle fails—soft modes scatter irregularly and don't obey ordinary dispersion. For curvature falling faster than $r^{ 3}$, extrapolation is well behaved (zero energy is regular). At the critical $r^{ 3}$ decay, geometry and dispersion balance and the rules change. Why it matters: This is where linearized gravity's low frequency physics becomes strongly long range correlated and cannot be treated perturbatively. Summary Latent geometry breaks extrapolation when: 1. Field space distance grows too large — new particles become light [AG 2024.07 681] 2. Discrete/singular limits dominate — path integrals become ambiguous [AG 2024.06 492, AG 2024.07 392] 3. Wavelengths are short compared to curvature scales — modes are suppressed [AG 2026.02 369] 4. Infrared decay is too slow — soft mode structure becomes singular [AG 2025.11 160] The common thread: extrapolation fails when the regime you're extrapolating into has fundamentally different physics from the regime you started in.