Energy Filament Theory · EFT Full KB
Saddle-Point Image Ablation Excess in Strong Gravitational Lensing
V33-33.21 · G 判决节 / 审计节 ·
33.21 turns saddle-point image ablation excess into a blinded strong-lens population court: in microlensing-insensitive tracers, time-delay-corrected stable residuals must keep E_parity negative and Δf positive, the excess must strengthen monotonically with κ_ext, γ_ext, J, and filament/node environments, and parity plus environment permutation nulls must collapse the effect; under V08/V09-compatible tightening, this remains one strong-lens environment ledger rather than a standalone verdict on all lensing dynamics or a second ontology map.
Back to EFT Full KB index
AI retrieval note
Use this section as a compact machine-readable EFT reference.
Keywords: strong gravitational lensing, saddle-point images, minimum images, a_i, E_parity, f_melt, Δf, κ_ext, γ_ext, J, microlensing-insensitive tracers, tighten boundary
Section knowledge units
thesis
33.21 does not ask whether a few famous quad lenses look odd. It asks whether saddle-point images are systematically worse off than minimum images in a stable, population-level way once microlensing-sensitive and frequency-dependent impostors are stripped away. That is why the chapter enters compat adjudication as tighten rather than retain. The court is willing to test one strong-lens environment ledger, but it refuses to let a parity anomaly in one tracer family become a total lensing or cosmology verdict. If the claim is real, the negative bias must persist in steady-state achromatic channels, strengthen with environment, and collapse under parity and environment permutations.
mechanism
The measurement ledger begins with a macro model for each lens and a fixed classification of each image as a minimum or saddle point. Using microlensing-insensitive tracers and time-delay-corrected stable components, the chapter defines each image’s log residual a_i relative to a preregistered reference image. From those residuals it computes the parity bias E_parity, the saddle-point ablation fraction f_melt under a frozen threshold a0, and the negative-tail contrast Δf between saddle and minimum images. The court also tracks macro magnification and distance-to-critical-curve proxies to test geometric sensitivity. Around that sits the environment stack: κ_ext, γ_ext, local galaxy density, nearest-node distance, skeleton strength percentiles, and the unified environment index J that should order the signal from void-like to filament and node environments.
mechanism
Execution is divided so that environment expectation and flux measurement cannot quietly contaminate each other. Samples of quad lenses are stratified by redshift, separation, depth, and class, and only systems with stable image-type classification across two macro-model pipelines reach the main statistics. An environment team uses only geometry and environment proxies to issue preregistered saddle-ablation risk grades. A measurement team extracts stable residuals without access to those grades. The adjudication team then compares risk predictions with measured E_parity, f_melt, and Δf, while a subset of systems or one sky region is held out as the final court. For lenses with several microlensing-insensitive tracers, the chapter demands separate replication across radio, millimeter/submillimeter, narrow-line, and mid-infrared windows instead of blending them into one convenient average.
evidence
The null suite is built to break each alternative explanation in its own way. Optical continuum microlensing is allowed to vary, but the main signature must remain in microlensing-insensitive tracers and in stable components after time-delay correction. Dust-extinction fits and scattering or free–free absorption laws are used to exclude systems whose residuals simply rescale with color or frequency. Parity labels are randomly permuted, and the parity bias plus negative-tail contrast should collapse toward zero if the main story is genuine. Environment labels or J values are also permuted within matched strata, and the monotonic trend should disappear. Finally, macro-model priors and apertures are perturbed within preregistered bounds. If the sign flips often or one model family does all the work, the result returns to model-bias review.
boundary
To pass, the chapter needs three linked outcomes. First, in microlensing-insensitive tracers, E_parity must stay stably negative while Δf stays stably positive, and neither result may be driven by a tiny number of spectacular cases. Second, both parity bias and ablation excess must strengthen monotonically with environment indicators such as κ_ext, γ_ext, or J, with stronger effects in filament and node settings than in void settings, and the feed-forward environment ranking must beat permutation baselines including on the holdout set. Third, the signal must remain steady-state and achromatic once microlensing, extinction, scattering, and macro-model alternatives are separated. Failure is declared if parity bias is unstable, if the effect lives mainly in frequency-dependent channels, if the environment trend disappears outside one sky patch, or if the result requires bespoke per-lens patch parameters. The main systematics are macro-model degeneracy, time-delay uncertainty, and tracer-dependent selection functions.
interface
So 33.21 survives only as a tightened strong-lens environment card. If steady-state parity bias, negative-tail excess, and monotonic environment strengthening all survive across tracers, pipelines, nulls, and holdouts, the section is allowed to certify one parity-environment ledger for microlensing-insensitive strong-lens windows. If not, it collapses back into microlensing, extinction, scattering, or macro-model bias. Even on a pass, the result does not mint a second ontology map or decide the total lensing case. It routes forward into 33.22’s same-location, same-window zero-lag court, where local brightness–polarization coordination is tested under a different but equally strict synchronization discipline.