Memory-Burden Effect in Black Holes & Quantum Systems

• Memory-burden effect is the stabilization mechanism in quantum many-body systems where stored information raises energy barriers, suppressing decay and emissions.
• Its application to black holes reveals that gapless memory modes delay Hawking evaporation and alter ringdown characteristics.
• Recent models extend the effect to primordial black holes and machine learning, offering new insights into critical dynamics and information storage.

The memory-burden effect denotes a class of phenomena in which stored information backreacts on the dynamics of the system that stores it, thereby suppressing decay, emission, or further state change. In its original and most developed sense, it is a quantum many-body effect in systems with enhanced memory-storage capacity: near a critical point, many memory modes become gapless, patterns can be stored at negligible energy cost, and subsequent departure from criticality makes those patterns energetically expensive, stabilizing the state that carries them (Dvali, 2018). This mechanism has been applied to black holes, where it is used to argue for deviations from semiclassical Hawking evaporation and for information-load–dependent modifications of classical perturbations, including black-hole spectroscopy and ringdown (Dvali et al., 2020). Later literature extended the term to phenomenological models of primordial black holes, gravitational-wave observables, and, in a distinct operational sense, to memory architectures in machine learning (Alexandre et al., 2024).

1. Microscopic definition and assisted gaplessness

The foundational formulation introduces the memory-burden effect as the dynamical stabilization of a holographic state by the load of information stored in its emergent gapless modes. The central statement is that among degenerate microstates, the ones with heavier loaded memories survive longer than those that store emptier patterns, because moving away from the critical state raises the energy cost of the stored pattern (Dvali, 2018).

In the microscopic model, a bosonic field inhabiting a $d$-dimensional sphere experiences a momentum-dependent attractive interaction. In the double-scaling regime,

$$N \rightarrow \infty\,, ~~ \Lambda \rightarrow \infty,\qquad \frac{e}{\Lambda} N^{\frac{3d}{3d -1}} = {\rm finite}\,,$$

a macroscopically occupied soft mode induces assisted gaplessness in a set of memory modes. The effective single-particle gaps become

$$\epsilon_k= e\omega_k \, \left(1 - \frac{N e}{\Lambda} \omega_k\right )^2 ,$$

and levels satisfying

$N k_d(k_d+d-1) = \frac{\Lambda}{e}$

become gapless (Dvali, 2018).

This criticality produces an exponentially large memory space. The basis states are

${N-n}_{a_0}\otimes {pattern}_{k} \, , \qquad {pattern}_{k} \equiv {n_{1}^{(k)} ,....,n^{(k)}_{\mathcal N_k} ,$

with memory-space dimension $2^{\mathcal N_k}$ for binary occupations. The total memory load is

$$n^{(k)} \equiv \sum_{ j= 1}^{\mathcal N_k} n_{j}^{(k)} ,$$

and the entropy obeys an area law,

$S = {\mathcal N}_k {\rm ln}2.$

The number of gapless modes scales as the area of a $(d-1)$-dimensional sphere,

$${\mathcal N}_k = \left ( \frac{R}{L_k} \right )^{d-1},$$

which is the microscopic origin of the holographic interpretation (Dvali, 2018).

The burden appears when the system is perturbed away from criticality. The gap shift and pattern energy cost are

$$N \rightarrow \infty\,, ~~ \Lambda \rightarrow \infty,\qquad \frac{e}{\Lambda} N^{\frac{3d}{3d -1}} = {\rm finite}\,,$$0

$$N \rightarrow \infty\,, ~~ \Lambda \rightarrow \infty,\qquad \frac{e}{\Lambda} N^{\frac{3d}{3d -1}} = {\rm finite}\,,$$1

Thus the energetic penalty grows with the memory load $$N \rightarrow \infty\,, ~~ \Lambda \rightarrow \infty,\qquad \frac{e}{\Lambda} N^{\frac{3d}{3d -1}} = {\rm finite}\,,$$2, and the system becomes harder to deplete as more information is stored (Dvali, 2018).

The time-evolution formula makes the stabilization explicit: $$N \rightarrow \infty\,, ~~ \Lambda \rightarrow \infty,\qquad \frac{e}{\Lambda} N^{\frac{3d}{3d -1}} = {\rm finite}\,,$$3 with

$$N \rightarrow \infty\,, ~~ \Lambda \rightarrow \infty,\qquad \frac{e}{\Lambda} N^{\frac{3d}{3d -1}} = {\rm finite}\,,$$4

Larger memory load reduces $$N \rightarrow \infty\,, ~~ \Lambda \rightarrow \infty,\qquad \frac{e}{\Lambda} N^{\frac{3d}{3d -1}} = {\rm finite}\,,$$5, suppressing oscillations and stabilizing the state (Dvali, 2018).

2. Black holes, saturons, and stabilization by stored information

The effect was subsequently embedded in a broader framework for black holes and saturons. In that formulation, systems of enhanced information capacity possess an exponentially large number of microstates within small energy gaps, and black holes saturate the corresponding entropy bounds (Dvali et al., 2024). The paper identifies a master mode $$N \rightarrow \infty\,, ~~ \Lambda \rightarrow \infty,\qquad \frac{e}{\Lambda} N^{\frac{3d}{3d -1}} = {\rm finite}\,,$$6 and memory modes $$N \rightarrow \infty\,, ~~ \Lambda \rightarrow \infty,\qquad \frac{e}{\Lambda} N^{\frac{3d}{3d -1}} = {\rm finite}\,,$$7, with a prototype Hamiltonian

$$N \rightarrow \infty\,, ~~ \Lambda \rightarrow \infty,\qquad \frac{e}{\Lambda} N^{\frac{3d}{3d -1}} = {\rm finite}\,,$$8

The effective memory-mode gap is

$$N \rightarrow \infty\,, ~~ \Lambda \rightarrow \infty,\qquad \frac{e}{\Lambda} N^{\frac{3d}{3d -1}} = {\rm finite}\,,$$9

Near $$\epsilon_k= e\omega_k \, \left(1 - \frac{N e}{\Lambda} \omega_k\right )^2 ,$$0, the modes become gapless and information is stored nearly for free; when $$\epsilon_k= e\omega_k \, \left(1 - \frac{N e}{\Lambda} \omega_k\right )^2 ,$$1 decreases, the gaps reopen and the stored pattern becomes expensive (Dvali et al., 2024).

For black holes, the proposed dictionary is that the master mode corresponds to soft constituent gravitons, while the memory modes correspond to hard, near-Planckian holographic degrees of freedom. The black-hole entropy is

$$\epsilon_k= e\omega_k \, \left(1 - \frac{N e}{\Lambda} \omega_k\right )^2 ,$$2

and the occupation number satisfies $$\epsilon_k= e\omega_k \, \left(1 - \frac{N e}{\Lambda} \omega_k\right )^2 ,$$3. The onset of the memory burden is estimated to occur at latest after an order-one fraction of the master occupation is lost, with an upper-bound time

$$\epsilon_k= e\omega_k \, \left(1 - \frac{N e}{\Lambda} \omega_k\right )^2 ,$$4

that is, at latest by the age of a black hole that has emitted about half its mass (Dvali et al., 2024).

The same framework distinguishes stabilization by quantum memory burden from stabilization by classical hair. Classical electric charge or spin stabilizes through long-range fields detectable at infinity, whereas the memory-burden effect does not require long-range hair and applies to neutral, non-spinning objects as well (Dvali et al., 2024). This distinction is sharpened in the later black-hole metamorphosis analysis, which argues that self-similar Hawking evaporation must drastically change no later than half-decay, because the stored information cannot be off-loaded to radiation efficiently (Dvali et al., 2020).

That metamorphosis paper formulates the memory burden through a master mode $$\epsilon_k= e\omega_k \, \left(1 - \frac{N e}{\Lambda} \omega_k\right )^2 ,$$5, memory modes $$\epsilon_k= e\omega_k \, \left(1 - \frac{N e}{\Lambda} \omega_k\right )^2 ,$$6, and an auxiliary out mode $$\epsilon_k= e\omega_k \, \left(1 - \frac{N e}{\Lambda} \omega_k\right )^2 ,$$7, with Hamiltonian

$$\epsilon_k= e\omega_k \, \left(1 - \frac{N e}{\Lambda} \omega_k\right )^2 ,$$8

The memory-burden parameter is

$$\epsilon_k= e\omega_k \, \left(1 - \frac{N e}{\Lambda} \omega_k\right )^2 ,$$9

For $N k_d(k_d+d-1) = \frac{\Lambda}{e}$0, the amplitude of $N k_d(k_d+d-1) = \frac{\Lambda}{e}$1 oscillations is suppressed by $N k_d(k_d+d-1) = \frac{\Lambda}{e}$2, so the stored pattern ties the master mode to criticality. The paper argues that after the metamorphosis point the emission rate is entropy suppressed,

$N k_d(k_d+d-1) = \frac{\Lambda}{e}$3

and the lifetime obeys

$N k_d(k_d+d-1) = \frac{\Lambda}{e}$4

with $N k_d(k_d+d-1) = \frac{\Lambda}{e}$5 representing additional entropy suppression beyond the semiclassical rate (Dvali et al., 2020).

3. Primordial black holes and memory-burden–modified evaporation

A large body of later work applies the effect to primordial black holes (PBHs), usually through phenomenological modifications of the Hawking mass-loss rate. A common baseline is the semiclassical evaporation law

$N k_d(k_d+d-1) = \frac{\Lambda}{e}$6

or, in alternative conventions,

$N k_d(k_d+d-1) = \frac{\Lambda}{e}$7

with $N k_d(k_d+d-1) = \frac{\Lambda}{e}$8 and lifetime scaling $N k_d(k_d+d-1) = \frac{\Lambda}{e}$9 in the absence of memory burden (Du et al., 19 May 2026).

A widely used memory-burden parameterization suppresses the semiclassical rate by an entropy-dependent factor,

${N-n}_{a_0}\otimes {pattern}_{k} \, , \qquad {pattern}_{k} \equiv {n_{1}^{(k)} ,....,n^{(k)}_{\mathcal N_k} ,$0

with ${N-n}_{a_0}\otimes {pattern}_{k} \, , \qquad {pattern}_{k} \equiv {n_{1}^{(k)} ,....,n^{(k)}_{\mathcal N_k} ,$1, often together with a “swift” onset after the PBH loses a fraction of its initial mass,

${N-n}_{a_0}\otimes {pattern}_{k} \, , \qquad {pattern}_{k} \equiv {n_{1}^{(k)} ,....,n^{(k)}_{\mathcal N_k} ,$2

In that implementation the MB regime begins after half-decay, and for ${N-n}_{a_0}\otimes {pattern}_{k} \, , \qquad {pattern}_{k} \equiv {n_{1}^{(k)} ,....,n^{(k)}_{\mathcal N_k} ,$3 the lifetime scaling changes from

${N-n}_{a_0}\otimes {pattern}_{k} \, , \qquad {pattern}_{k} \equiv {n_{1}^{(k)} ,....,n^{(k)}_{\mathcal N_k} ,$4

to

${N-n}_{a_0}\otimes {pattern}_{k} \, , \qquad {pattern}_{k} \equiv {n_{1}^{(k)} ,....,n^{(k)}_{\mathcal N_k} ,$5

so ${N-n}_{a_0}\otimes {pattern}_{k} \, , \qquad {pattern}_{k} \equiv {n_{1}^{(k)} ,....,n^{(k)}_{\mathcal N_k} ,$6 when ${N-n}_{a_0}\otimes {pattern}_{k} \, , \qquad {pattern}_{k} \equiv {n_{1}^{(k)} ,....,n^{(k)}_{\mathcal N_k} ,$7 (Du et al., 19 May 2026).

This framework has been combined with regular, non-singular PBH metrics. For the Hayward, Bardeen, and Simpson–Visser black holes, the regularization lowers the Hawking temperature relative to Schwarzschild, and MB suppression further reduces evaporation after ${N-n}_{a_0}\otimes {pattern}_{k} \, , \qquad {pattern}_{k} \equiv {n_{1}^{(k)} ,....,n^{(k)}_{\mathcal N_k} ,$8. For the benchmark ${N-n}_{a_0}\otimes {pattern}_{k} \, , \qquad {pattern}_{k} \equiv {n_{1}^{(k)} ,....,n^{(k)}_{\mathcal N_k} ,$9 and $2^{\mathcal N_k}$0, a new fully open PBH dark-matter window was reported: $2^{\mathcal N_k}$1

$2^{\mathcal N_k}$2

$2^{\mathcal N_k}$3

with energy-injection ratios

$2^{\mathcal N_k}$4

under the assumptions used there (Du et al., 19 May 2026).

Another line of work emphasizes that the cosmological implications depend sensitively on how the transition into the burdened regime is modeled. A sharp two-stage prescription can open a light-mass PBH dark-matter window, whereas a continuous transition can keep Hawking emission active during Big Bang nucleosynthesis and recombination. One phenomenological smooth model writes

$2^{\mathcal N_k}$5

with

$2^{\mathcal N_k}$6

In that analysis, PBHs lighter than about $2^{\mathcal N_k}$7 cannot constitute all or most of the dark matter if the transition is continuous, because BBN and CMB bounds close the would-be window (Montefalcone et al., 26 Mar 2025).

A related study replaced the instantaneous transition by a continuous tanh crossover and compared additive and multiplicative combinations of the two rates. The multiplicative prescription produced tighter BBN exclusion curves than the additive one, and both were tighter than the instantaneous transition. Over

$2^{\mathcal N_k}$8

the additive case can permit

$2^{\mathcal N_k}$9

where the multiplicative case gives

$$n^{(k)} \equiv \sum_{ j= 1}^{\mathcal N_k} n_{j}^{(k)} ,$$0

showing that the rate-combination rule is itself a material assumption (Zhang et al., 3 Jun 2026).

4. High-energy messengers, gravitational waves, and hot-spot phenomenology

Beyond abundance constraints, the memory-burden effect has been used to predict modified astrophysical signatures of evaporating PBHs. One neutrino-focused approach implements an energy-dependent deformation of the Hawking spectrum,

$$n^{(k)} \equiv \sum_{ j= 1}^{\mathcal N_k} n_{j}^{(k)} ,$$1

so that

$$n^{(k)} \equiv \sum_{ j= 1}^{\mathcal N_k} n_{j}^{(k)} ,$$2

This leaves the infrared regime approximately unchanged while suppressing the ultraviolet tail, reducing the total luminosity by a mass-independent factor

$$n^{(k)} \equiv \sum_{ j= 1}^{\mathcal N_k} n_{j}^{(k)} ,$$3

The modified lifetime becomes

$$n^{(k)} \equiv \sum_{ j= 1}^{\mathcal N_k} n_{j}^{(k)} ,$$4

For $$n^{(k)} \equiv \sum_{ j= 1}^{\mathcal N_k} n_{j}^{(k)} ,$$5, the paper reports $$n^{(k)} \equiv \sum_{ j= 1}^{\mathcal N_k} n_{j}^{(k)} ,$$6, implying $$n^{(k)} \equiv \sum_{ j= 1}^{\mathcal N_k} n_{j}^{(k)} ,$$7, and shows that IceCube constraints on $$n^{(k)} \equiv \sum_{ j= 1}^{\mathcal N_k} n_{j}^{(k)} ,$$8 are weakened by factors of about $$n^{(k)} \equiv \sum_{ j= 1}^{\mathcal N_k} n_{j}^{(k)} ,$$9–$S = {\mathcal N}_k {\rm ln}2.$0 at $S = {\mathcal N}_k {\rm ln}2.$1 (Chaudhuri, 8 Apr 2026).

A different neutrino analysis models the burdened phase by

$S = {\mathcal N}_k {\rm ln}2.$2

with onset at $S = {\mathcal N}_k {\rm ln}2.$3, $S = {\mathcal N}_k {\rm ln}2.$4. Using IceCube HESE and EHE data, it finds that neutrino constraints are stronger than gamma-ray bounds at large $S = {\mathcal N}_k {\rm ln}2.$5, and that future facilities such as IceCube-Gen2 and GRAND can probe scenarios with highly suppressed evaporation and very light PBHs (Chianese et al., 2024).

The same general framework has been extended to induced stochastic gravitational-wave backgrounds. In one study the MB effect on Hawking evaporation of ultra-low-mass PBHs was shown to mimic the signature of a non-standard reheating epoch before PBH domination in the secondary SGWB sourced by PBH density fluctuations. The degeneracy can be broken if the inflationary first peak in the SGWB is also detected, because that peak depends strongly on the pre-PBH equation of state, whereas the PBH-induced peak can be reproduced by MB-induced shifts of $S = {\mathcal N}_k {\rm ln}2.$6, $S = {\mathcal N}_k {\rm ln}2.$7, $S = {\mathcal N}_k {\rm ln}2.$8, and $S = {\mathcal N}_k {\rm ln}2.$9 (Bhaumik et al., 2024).

Another induced-GW analysis focused on PBH dark matter with memory burden and reported the characteristic prediction

$(d-1)$0

arguing that PBH dark matter with initial mass greater than about $(d-1)$1 grams can be tested by future observations such as Cosmic Explorer (Kohri et al., 2024).

The effect has also been propagated into plasma-heating phenomenology around evaporating PBHs. In a transfer-function approach, MB modifies both the mass-loss rate and Hawking temperature,

$(d-1)$2

leading to simple scaling laws for hot-spot core quantities,

$(d-1)$3

In the “vanilla” rigid MB scenario, where the evaporation rate and Hawking temperature freeze below $(d-1)$4, the hot-spot temperature is substantially lowered, and hot spots form only if

$(d-1)$5

under the assumptions used in that paper (Levy et al., 21 Nov 2025).

5. Swift memory burden and black-hole spectroscopy

A distinct but related development concerns how information load affects classical perturbations of already formed black holes. This “swift memory burden” effect argues that black holes with different information loads, although degenerate in the ground state, respond differently to perturbations. The controlling quantity is the memory-burden parameter

$(d-1)$6

which measures the fraction of the black hole’s memory space occupied by the information load (Dvali, 26 Sep 2025).

In the prototype merger framework, the Hamiltonian takes the form

$(d-1)$7

with the black-hole dictionary

$(d-1)$8

The critical fractional loss that triggers strong swift burden is

$(d-1)$9

For $${\mathcal N}_k = \left ( \frac{R}{L_k} \right )^{d-1},$$0, this gives $${\mathcal N}_k = \left ( \frac{R}{L_k} \right )^{d-1},$$1, while for $${\mathcal N}_k = \left ( \frac{R}{L_k} \right )^{d-1},$$2, $${\mathcal N}_k = \left ( \frac{R}{L_k} \right )^{d-1},$$3 (Dvali, 26 Sep 2025).

The framework predicts a burden-induced suppression of radiation intensity into a frequency bin $${\mathcal N}_k = \left ( \frac{R}{L_k} \right )^{d-1},$$4,

$${\mathcal N}_k = \left ( \frac{R}{L_k} \right )^{d-1},$$5

and a prompt onset time for the burden under perturbation,

$${\mathcal N}_k = \left ( \frac{R}{L_k} \right )^{d-1},$$6

The paper proposes that this should shift quasi-normal-mode frequencies and suppress damping in black-hole mergers, with parametric scalings

$${\mathcal N}_k = \left ( \frac{R}{L_k} \right )^{d-1},$$7

The effect is therefore load dependent: two black holes with the same $${\mathcal N}_k = \left ( \frac{R}{L_k} \right )^{d-1},$$8 but different $${\mathcal N}_k = \left ( \frac{R}{L_k} \right )^{d-1},$$9 are predicted to have different ringdown responses (Dvali, 26 Sep 2025).

This proposal was then connected to gravitational-wave data through a minimal phenomenological framework for black-hole spectroscopy. The ringdown spectrum is modeled as a Lorentzian

$$N \rightarrow \infty\,, ~~ \Lambda \rightarrow \infty,\qquad \frac{e}{\Lambda} N^{\frac{3d}{3d -1}} = {\rm finite}\,,$$00

multiplied by an SMB suppression factor

$$N \rightarrow \infty\,, ~~ \Lambda \rightarrow \infty,\qquad \frac{e}{\Lambda} N^{\frac{3d}{3d -1}} = {\rm finite}\,,$$01

For $$N \rightarrow \infty\,, ~~ \Lambda \rightarrow \infty,\qquad \frac{e}{\Lambda} N^{\frac{3d}{3d -1}} = {\rm finite}\,,$$02, the peak is redshifted to

$$N \rightarrow \infty\,, ~~ \Lambda \rightarrow \infty,\qquad \frac{e}{\Lambda} N^{\frac{3d}{3d -1}} = {\rm finite}\,,$$03

equivalently,

$$N \rightarrow \infty\,, ~~ \Lambda \rightarrow \infty,\qquad \frac{e}{\Lambda} N^{\frac{3d}{3d -1}} = {\rm finite}\,,$$04

Notably, this peak shift is independent of $$N \rightarrow \infty\,, ~~ \Lambda \rightarrow \infty,\qquad \frac{e}{\Lambda} N^{\frac{3d}{3d -1}} = {\rm finite}\,,$$05, so frequency-shift measurements constrain $$N \rightarrow \infty\,, ~~ \Lambda \rightarrow \infty,\qquad \frac{e}{\Lambda} N^{\frac{3d}{3d -1}} = {\rm finite}\,,$$06 but not $$N \rightarrow \infty\,, ~~ \Lambda \rightarrow \infty,\qquad \frac{e}{\Lambda} N^{\frac{3d}{3d -1}} = {\rm finite}\,,$$07 (Yuan et al., 22 Oct 2025).

Using LVK posteriors for the $$N \rightarrow \infty\,, ~~ \Lambda \rightarrow \infty,\qquad \frac{e}{\Lambda} N^{\frac{3d}{3d -1}} = {\rm finite}\,,$$08 and $$N \rightarrow \infty\,, ~~ \Lambda \rightarrow \infty,\qquad \frac{e}{\Lambda} N^{\frac{3d}{3d -1}} = {\rm finite}\,,$$09 QNM frequencies of GW250114, that analysis found a lower bound

$$N \rightarrow \infty\,, ~~ \Lambda \rightarrow \infty,\qquad \frac{e}{\Lambda} N^{\frac{3d}{3d -1}} = {\rm finite}\,,$$10

with the constraint dominated by the $$N \rightarrow \infty\,, ~~ \Lambda \rightarrow \infty,\qquad \frac{e}{\Lambda} N^{\frac{3d}{3d -1}} = {\rm finite}\,,$$11 mode. A Fisher forecast for a GW250114-like source observed by Cosmic Explorer gave

$$N \rightarrow \infty\,, ~~ \Lambda \rightarrow \infty,\qquad \frac{e}{\Lambda} N^{\frac{3d}{3d -1}} = {\rm finite}\,,$$12

disfavoring rapid gap reopening and indicating that next-generation spectroscopy can tighten constraints by roughly three orders of magnitude (Yuan et al., 22 Oct 2025).

6. Conceptual extensions, alternative usages, and open distinctions

The term “memory burden” has also been reused outside black-hole physics, but the meaning is not uniform. In generative retrieval, “memory can be a burden” refers to the limitations of memorizing the corpus in model parameters: poor fine-grained memory accuracy, growing confusion with corpus size, and high update costs. The proposed remedy is a coarse-to-fine retrieval architecture in which only cluster-level semantics are memorized generatively, while dense retrieval handles fine-grained matching (Yuan et al., 2024).

In conversational AI, the “Memory-Burden Effect” names the inference-time burden created when a dialogue model must perform cross-session synthesis, temporal reasoning, contradiction resolution, and causal linking at response time. PREMem addresses this by shifting reasoning from inference-time to pre-storage memory construction, using memory fragments, cross-session linking, and reasoning memories such as extension, accumulation, specification, transformation, and implication (Kim et al., 13 Sep 2025). This usage is operational rather than quantum-statistical.

Within physics itself, several distinctions remain active. One concerns whether the memory burden should be modeled as an instantaneous transition, a smooth crossover, or an energy-dependent spectral deformation. Another concerns whether the relevant observables are late-time evaporation products, hot-spot morphology, or classical perturbations and ringdown. A further distinction is between the original gradual burden associated with evaporation and the swift burden associated with coherent perturbations on merger timescales. This suggests that “memory burden” functions as a family resemblance term rather than a single settled formalism.

A plausible implication is that the most robust common core lies in the relation between enhanced memory capacity and dynamical inhibition of decay or response. In the original microscopic papers, the burden arises because information stored at negligible energy cost near criticality becomes expensive away from criticality (Dvali, 2018). In black-hole evaporation models, this is implemented as entropy-powered suppression of Hawking emission or of high-energy spectral tails (Du et al., 19 May 2026). In swift-memory-burden ringdown models, the same logic is transferred to classical perturbations, yielding mode-dependent detuning and suppressed radiation near the unburdened resonance (Yuan et al., 22 Oct 2025).

The principal controversy is therefore not whether different communities use the term, but which effective implementation correctly captures the underlying microscopic mechanism in black holes. The existing literature shows that quantitative conclusions can change materially with the onset criterion, crossover width, suppression law, and observable chosen. That makes the memory-burden effect less a single finished theory than a developing research program linking information storage, criticality, and measurable deviations from standard dynamics.