Electroweak SNR: Theory & Cosmological Impact

• Electroweak SNR is characterized by a nonzero Higgs order parameter at high temperatures, defying the Standard Model expectation of symmetry restoration above ~160 GeV.
• Different models use scalar, fermionic, or extended Higgs sector mechanisms to induce a negative high-temperature thermal mass, enabling diverse phase histories.
• This phenomenon offers practical implications including sphaleron suppression, modified phase transitions, and distinct cosmological and collider signatures.

Electroweak symmetry non-restoration (EW SNR) denotes thermal histories in which the Higgs order parameter remains nonzero at high temperature, in contrast to the Standard Model expectation that finite-temperature corrections restore electroweak symmetry above approximately $160$ GeV. In the contemporary literature, EW SNR includes both cases where the broken phase persists to arbitrarily high temperature and cases with more intricate sequences, such as temporary restoration, two-step symmetry breaking in an extended Higgs sector, or metastable broken vacua that survive cosmological evolution despite a symmetric global minimum (Matsedonskyi et al., 2020, Meade et al., 2018, Biekötter et al., 2021).

1. Definition and thermal criteria

The finite-temperature analysis of EW SNR is typically formulated through a one-loop, daisy-improved effective potential,

$$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$

or model-specific extensions of this structure, with $J_B$ and $J_F$ thermal functions and Arnold–Espinosa resummation for infrared-sensitive bosonic modes. This framework is used across singlet extensions, two-Higgs-doublet models, the N2HDM, and scalar-condensate constructions (Chang et al., 2022, Biekötter et al., 2021, Aoki et al., 2023).

In the Standard Model, the relevant high-temperature effect is a positive Higgs thermal mass. One explicit formulation gives

$$\Delta m_h^2|_{\rm SM}(T)\simeq T^2\left[\frac{y_t^2}{4}+\frac{\lambda}{2}+\frac{3g^2}{16}+\frac{g'^2}{16}\right]\simeq 0.4\,T^2,$$

which drives the Higgs vacuum expectation value to zero for $T\gtrsim160$ GeV (Matsedonskyi et al., 2020). A closely related singlet-scalar analysis writes

$$\Pi_h(T)=T^2\left[\frac{\lambda_t^2}{4}+\frac{3g^2}{16}+\frac{g'^2}{16}+\frac{\lambda_h}{2}+N_s\frac{\lambda_{h\phi}}{12}\right],$$

so that EW SNR requires the negative portal contribution to overcome the positive Standard Model terms (Meade et al., 2018).

Different model classes encode the SNR condition in different effective coefficients. In singlet-scalar models with negative portal coupling one may define

$c_{\rm eff}=c_H+N\frac{\lambda_{HS}}{12},$

and require $c_{\rm eff}<0$ at high temperature (Baldes et al., 2018). In a fermionic construction with Higgs-dependent singlet-fermion masses,

$m_N(h)=m_N^0-\frac{\lambda_N}{\Lambda}h^2,$

the high-temperature criterion is $$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$0, together with $$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$1 (Matsedonskyi et al., 2020). In the scalar-condensate scenario, the Higgs thermal mass is written as

$$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$2

with SNR requiring

$$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$3

so that the condensate-induced negative contribution dominates the Standard Model plasma contribution (Chang et al., 2022).

A recurrent diagnostic is the ratio $$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$4 or its model-dependent generalization. In the fermionic SNR literature, $$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$5 is the criterion for sphaleron suppression in the broken phase (Matsedonskyi et al., 2020). In multi-Higgs models this can be replaced by a composite order parameter such as

$$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$6

or

$$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$7

depending on the field content (Matsedonskyi et al., 2021, Carena et al., 2021).

2. Dynamical mechanisms that realize SNR

A standard mechanism employs many new scalar degrees of freedom with negative Higgs-portal couplings. In the large-$$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$8 singlet constructions of the form

$$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$9

with $J_B$0, each singlet contributes a negative thermal correction to the Higgs mass, and for sufficiently large $J_B$1 one obtains either symmetry non-restoration or temporary restoration. A quoted analytic estimate is $J_B$2, while explicit examples with $J_B$3 display all three qualitative histories: symmetry restoration, temporary restoration, and SNR (Meade et al., 2018). Closely related high-scale baryogenesis constructions also use $J_B$4 real singlets $J_B$5 with $J_B$6, requiring

$J_B$7

for the total Higgs thermal mass coefficient to become negative (Baldes et al., 2018).

A distinct fermionic mechanism replaces negative scalar thermal fluctuations by singlet fermions whose masses decrease with $J_B$8. The finite-temperature potential develops a dip at the field value where $J_B$9, and the curvature at the origin becomes negative when the fermionic contribution dominates the Standard Model term. This yields several thermal histories: Standard Model-like restoration for $J_F$0, “high-T SNR only” when the fermions decouple at lower temperature, and “continuous SNR” when $J_F$1 so that no intermediate restored phase appears (Matsedonskyi et al., 2020).

In a two-Higgs-doublet realization connected to dark matter, the high-temperature broken direction is $J_F$2, not the Standard Model-like $J_F$3. The relevant thermal correction is a negative contribution to the $J_F$4 mass from singlet fermions $J_F$5 coupled through

$J_F$6

Defining $J_F$7, the model finds that

$J_F$8

and that temperatures up to $J_F$9TeV can be maintained with small multiplicities $$\Delta m_h^2|_{\rm SM}(T)\simeq T^2\left[\frac{y_t^2}{4}+\frac{\lambda}{2}+\frac{3g^2}{16}+\frac{g'^2}{16}\right]\simeq 0.4\,T^2,$$0–6 (Matsedonskyi et al., 2021).

A qualitatively different mechanism is provided by a conserved-charge-induced Bose–Einstein condensate. In the scalar-condensate model, the Standard Model is extended by a single complex scalar $$\Delta m_h^2|_{\rm SM}(T)\simeq T^2\left[\frac{y_t^2}{4}+\frac{\lambda}{2}+\frac{3g^2}{16}+\frac{g'^2}{16}\right]\simeq 0.4\,T^2,$$1 charged under a global $$\Delta m_h^2|_{\rm SM}(T)\simeq T^2\left[\frac{y_t^2}{4}+\frac{\lambda}{2}+\frac{3g^2}{16}+\frac{g'^2}{16}\right]\simeq 0.4\,T^2,$$2, with tree-level potential

$$\Delta m_h^2|_{\rm SM}(T)\simeq T^2\left[\frac{y_t^2}{4}+\frac{\lambda}{2}+\frac{3g^2}{16}+\frac{g'^2}{16}\right]\simeq 0.4\,T^2,$$3

A global-charge asymmetry $$\Delta m_h^2|_{\rm SM}(T)\simeq T^2\left[\frac{y_t^2}{4}+\frac{\lambda}{2}+\frac{3g^2}{16}+\frac{g'^2}{16}\right]\simeq 0.4\,T^2,$$4 can force a Bose–Einstein condensate once $$\Delta m_h^2|_{\rm SM}(T)\simeq T^2\left[\frac{y_t^2}{4}+\frac{\lambda}{2}+\frac{3g^2}{16}+\frac{g'^2}{16}\right]\simeq 0.4\,T^2,$$5. In the quartic-dominated regime the condensate amplitude obeys

$$\Delta m_h^2|_{\rm SM}(T)\simeq T^2\left[\frac{y_t^2}{4}+\frac{\lambda}{2}+\frac{3g^2}{16}+\frac{g'^2}{16}\right]\simeq 0.4\,T^2,$$6

and the SNR condition becomes

$$\Delta m_h^2|_{\rm SM}(T)\simeq T^2\left[\frac{y_t^2}{4}+\frac{\lambda}{2}+\frac{3g^2}{16}+\frac{g'^2}{16}\right]\simeq 0.4\,T^2,$$7

This construction is presented as a minimal benchmark model in which one complex scalar field with a sufficiently large global-charge asymmetry keeps electroweak symmetry broken up to temperatures well above the electroweak scale (Chang et al., 2022).

A further line of work uses an inert Higgs sector and a large singlet sector. In that setup, an inert phase $$\Delta m_h^2|_{\rm SM}(T)\simeq T^2\left[\frac{y_t^2}{4}+\frac{\lambda}{2}+\frac{3g^2}{16}+\frac{g'^2}{16}\right]\simeq 0.4\,T^2,$$8 with $$\Delta m_h^2|_{\rm SM}(T)\simeq T^2\left[\frac{y_t^2}{4}+\frac{\lambda}{2}+\frac{3g^2}{16}+\frac{g'^2}{16}\right]\simeq 0.4\,T^2,$$9, $T\gtrsim160$0 is non-restoring when $T\gtrsim160$1, typically because $T\gtrsim160$2 and $T\gtrsim160$3 dominates the positive contributions to the inert thermal mass. Explicit benchmark scenarios with $T\gtrsim160$4 and $T\gtrsim160$5 exhibit SNR up to $T\gtrsim160$6 GeV and $T\gtrsim160$7 GeV, respectively (Carena et al., 2021).

3. Vacuum structure, metastability, and thermal phase histories

EW SNR is not equivalent to a unique phase diagram. Even within simple singlet models, the thermal history may show full restoration, temporary restoration, or a broken phase at all temperatures. The “temporary restoration” regime is particularly important because it demonstrates that a negative high-temperature thermal mass does not by itself imply a monotonic evolution of the Higgs expectation value (Meade et al., 2018). The singlet-fermion model makes the same point in a different language by distinguishing “high-T SNR only” from “continuous SNR” (Matsedonskyi et al., 2020).

Extended Higgs sectors make the multiplicity of histories explicit. In the two-Higgs-doublet dark-matter model, numerical minimization of the full one-loop finite-temperature potential yields a two-step sequence: for $T\gtrsim160$8–300 GeV the system is in the $T\gtrsim160$9, $$\Pi_h(T)=T^2\left[\frac{\lambda_t^2}{4}+\frac{3g^2}{16}+\frac{g'^2}{16}+\frac{\lambda_h}{2}+N_s\frac{\lambda_{h\phi}}{12}\right],$$0 vacuum with $$\Pi_h(T)=T^2\left[\frac{\lambda_t^2}{4}+\frac{3g^2}{16}+\frac{g'^2}{16}+\frac{\lambda_h}{2}+N_s\frac{\lambda_{h\phi}}{12}\right],$$1, while at lower temperature it transitions to the conventional $$\Pi_h(T)=T^2\left[\frac{\lambda_t^2}{4}+\frac{3g^2}{16}+\frac{g'^2}{16}+\frac{\lambda_h}{2}+N_s\frac{\lambda_{h\phi}}{12}\right],$$2, $$\Pi_h(T)=T^2\left[\frac{\lambda_t^2}{4}+\frac{3g^2}{16}+\frac{g'^2}{16}+\frac{\lambda_h}{2}+N_s\frac{\lambda_{h\phi}}{12}\right],$$3 vacuum (Matsedonskyi et al., 2021). In the inert-doublet plus singlet construction, the quoted benchmarks pass through multi-step histories involving $$\Pi_h(T)=T^2\left[\frac{\lambda_t^2}{4}+\frac{3g^2}{16}+\frac{g'^2}{16}+\frac{\lambda_h}{2}+N_s\frac{\lambda_{h\phi}}{12}\right],$$4, $$\Pi_h(T)=T^2\left[\frac{\lambda_t^2}{4}+\frac{3g^2}{16}+\frac{g'^2}{16}+\frac{\lambda_h}{2}+N_s\frac{\lambda_{h\phi}}{12}\right],$$5, and $$\Pi_h(T)=T^2\left[\frac{\lambda_t^2}{4}+\frac{3g^2}{16}+\frac{g'^2}{16}+\frac{\lambda_h}{2}+N_s\frac{\lambda_{h\phi}}{12}\right],$$6, while maintaining $$\Pi_h(T)=T^2\left[\frac{\lambda_t^2}{4}+\frac{3g^2}{16}+\frac{g'^2}{16}+\frac{\lambda_h}{2}+N_s\frac{\lambda_{h\phi}}{12}\right],$$7 throughout (Carena et al., 2021).

The relation between SNR and vacuum structure can be even more indirect. In the “Global Electroweak Symmetric Vacuum” model, the $$\Pi_h(T)=T^2\left[\frac{\lambda_t^2}{4}+\frac{3g^2}{16}+\frac{g'^2}{16}+\frac{\lambda_h}{2}+N_s\frac{\lambda_{h\phi}}{12}\right],$$8 symmetric point $$\Pi_h(T)=T^2\left[\frac{\lambda_t^2}{4}+\frac{3g^2}{16}+\frac{g'^2}{16}+\frac{\lambda_h}{2}+N_s\frac{\lambda_{h\phi}}{12}\right],$$9 is the global minimum once

$c_{\rm eff}=c_H+N\frac{\lambda_{HS}}{12},$0

while the electroweak-breaking vacuum at $c_{\rm eff}=c_H+N\frac{\lambda_{HS}}{12},$1 remains metastable and long-lived. Imposing vacuum-structure, quantum-tunneling, and thermal-tunneling constraints yields the quoted “safe” window

$c_{\rm eff}=c_H+N\frac{\lambda_{HS}}{12},$2

together with $c_{\rm eff}=c_H+N\frac{\lambda_{HS}}{12},$3 and $c_{\rm eff}=c_H+N\frac{\lambda_{HS}}{12},$4 (Bai et al., 2021). This construction shows that an early universe with EW SNR can terminate in the present electroweak vacuum even when that vacuum is not the global minimum.

The N2HDM sharpens the distinction between local curvature conditions and physically realized transitions. In that framework, $c_{\rm eff}=c_H+N\frac{\lambda_{HS}}{12},$5 with $c_{\rm eff}=c_H+N\frac{\lambda_{HS}}{12},$6 implies that the electroweak symmetry remains broken up to arbitrarily high temperature, whereas a negative $c_{\rm eff}=c_H+N\frac{\lambda_{HS}}{12},$7 can destabilize only the singlet direction and still leave electroweak symmetry restored in the stable minima. The analysis also emphasizes that “the existence of a critical temperature at which the electroweak phase becomes the deepest minimum is not sufficient for a transition to take place,” so the tunnelling probability to the electroweak minimum must be computed explicitly (Biekötter et al., 2021).

The 2HDM Type I provides an additional example in which SNR coexists with an intermediate charge-breaking phase. The one-loop thermal analysis identifies viable points with a charge-breaking phase at intermediate temperature and non-restoration at high temperature, whereas points with restoration are excluded because the charge-breaking phase then forces $c_{\rm eff}=c_H+N\frac{\lambda_{HS}}{12},$8 GeV (Aoki et al., 2023).

4. Representative realizations and quoted parameter regimes

The literature supports a useful classification by the dynamical origin of the negative high-temperature Higgs curvature. The following representative regimes are explicitly quoted.

Framework	SNR trigger	Quoted regime
$c_{\rm eff}=c_H+N\frac{\lambda_{HS}}{12},$9 singlet scalars (Meade et al., 2018)	Negative portal $c_{\rm eff}<0$0 with large $c_{\rm eff}<0$1	$c_{\rm eff}<0$2
High-scale singlet scalars (Baldes et al., 2018)	$c_{\rm eff}<0$3 real singlets $c_{\rm eff}<0$4 with $c_{\rm eff}<0$5	$c_{\rm eff}<0$6
One complex scalar condensate (Chang et al., 2022)	$c_{\rm eff}<0$7 asymmetry and Bose–Einstein condensate	$c_{\rm eff}<0$8; SNR up to $c_{\rm eff}<0$9 GeV or higher
Singlet fermions (Matsedonskyi et al., 2020)	Higgs-dependent fermion mass $m_N(h)=m_N^0-\frac{\lambda_N}{\Lambda}h^2,$0	$m_N(h)=m_N^0-\frac{\lambda_N}{\Lambda}h^2,$1
2HDM + singlet-fermion dark matter (Matsedonskyi et al., 2021)	Negative $m_N(h)=m_N^0-\frac{\lambda_N}{\Lambda}h^2,$2 thermal mass from $m_N(h)=m_N^0-\frac{\lambda_N}{\Lambda}h^2,$3	$m_N(h)=m_N^0-\frac{\lambda_N}{\Lambda}h^2,$4–6, $m_N(h)=m_N^0-\frac{\lambda_N}{\Lambda}h^2,$5, $m_N(h)=m_N^0-\frac{\lambda_N}{\Lambda}h^2,$6 TeV
Inert Higgs + singlets (Carena et al., 2021)	$m_N(h)=m_N^0-\frac{\lambda_N}{\Lambda}h^2,$7 from $m_N(h)=m_N^0-\frac{\lambda_N}{\Lambda}h^2,$8 and large $m_N(h)=m_N^0-\frac{\lambda_N}{\Lambda}h^2,$9	$$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$00 or $$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$01; SNR to $$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$02 GeV or $$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$03 GeV

These examples also illustrate how “minimality” depends on the trigger. Classic scalar SNR models often invoke $$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$04 additional scalars, whereas the scalar-condensate construction argues that one complex scalar field can suffice if it carries a sufficiently large conserved asymmetry (Chang et al., 2022). By contrast, fermionic and dark-matter realizations obtain SNR with small multiplicities, but typically only up to $$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$05TeV temperatures rather than the much higher scales quoted in large-$$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$06 inert-sector models (Matsedonskyi et al., 2021, Carena et al., 2021).

5. Cosmological consequences

The central cosmological consequence of EW SNR is sphaleron suppression in a phase with nonzero electroweak breaking. The standard condition is

$$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$07

which ensures that electroweak sphaleron transitions are exponentially suppressed and a previously generated baryon asymmetry is not washed out (Matsedonskyi et al., 2020). This basic mechanism is realized in several concrete ways: in the two-Higgs-doublet dark-matter model the combination $$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$08 remains larger than $$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$09 down to $$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$10 GeV (Matsedonskyi et al., 2021), while in the inert-doublet plus singlet benchmarks the washout dilution is negligible ($$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$11) for central $$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$12, and remains $$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$13 even with $$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$14 (Carena et al., 2021).

Because sphaleron washout is suppressed already at $$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$15, EW SNR broadens the range of viable baryogenesis mechanisms. One high-scale scenario links EW SNR to a high-scale electroweak phase transition and flavor-dependent CP violation, with the broken phase persisting after the transition so that the baryon asymmetry is not erased (Baldes et al., 2018). The fermionic SNR framework makes the same point in more general terms, noting that high-temperature first-order transitions in other sectors can generate baryon asymmetry while the always-broken electroweak phase protects it (Matsedonskyi et al., 2020). In the large-$$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$16 singlet model, the absence of sphaleron reprocessing implies that “standard EW baryogenesis and sphaleron reprocessing of L-asymmetry are inactive,” which instead invites high-scale mechanisms such as GUT or Affleck–Dine baryogenesis (Meade et al., 2018).

Gravitational-wave phenomenology depends on whether SNR is accompanied by additional first-order transitions. Pure SNR or temporary restoration in the minimal singlet model leads to either no electroweak phase transition or only second-order transitions, and hence no stochastic gravitational-wave background from bubble collisions (Meade et al., 2018). By contrast, the N2HDM admits strong first-order transitions with stochastic backgrounds in the LISA band for $$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$17–200 GeV (Biekötter et al., 2021). An even more elaborate possibility appears when EW SNR induces a large Higgs VEV that triggers color breaking through a negative Higgs–triplet quartic. In that setup, both the color-breaking and color-restoration transitions are first order, with benchmark values $$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$18, $$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$19, and $$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$20 and $$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$21, producing peaks in the DECIGO/BBO band (Chao et al., 2021).

Recent work also embeds SNR into broader hidden-sector cosmologies. In a supersymmetric Twin Higgs framework, SNR below the Twin electroweak scale ($$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$22TeV) is driven by mirror symmetry breaking in the Yukawa couplings, and the construction is combined with right-handed neutrinos to reduce dark relativistic degrees of freedom to a level consistent with CMB constraints, while also permitting a connection to minimal axiogenesis (Badziak et al., 21 Aug 2025).

6. Phenomenology, limitations, and major points of debate

The collider and low-energy implications of EW SNR vary sharply across models. In the metastable “Global Electroweak Symmetric Vacuum” construction, the same couplings that produce SNR imply substantial modifications of the Higgs self-interactions: $$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$23 The model also predicts sizable off-shell Higgs invisible-decay signals, with a current $$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$24 CL bound $$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$25 for $$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$26 GeV and an HL-LHC projection $$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$27 (Bai et al., 2021).

The dark-matter-motivated two-Higgs-doublet realization predicts a spin-independent nucleon cross section $$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$28–$$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$29 for $$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$30–500 GeV, just below current XENON1T/LUX/PandaX bounds but within reach of XENONnT, LZ, and DARWIN. It also permits exotic Higgs decays $$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$31 when $$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$32, and direct searches for $$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$33 in the $$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$34–150 GeV range (Matsedonskyi et al., 2021). In the N2HDM, collider signatures include the channels $$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$35 and $$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$36, with $$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$37 pb at 13 TeV in consistent first-order-transition and SNR regions (Biekötter et al., 2021). In the 2HDM Type I, the surviving charge-breaking plus SNR region is characterized by $$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$38 GeV, $$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$39, $$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$40, and $$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$41 (Aoki et al., 2023).

At the same time, some SNR constructions can be made nearly invisible experimentally. In the large-$$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$42 singlet model, deviations in Higgs observables scale as powers of $$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$43 divided by $$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$44, so “SNR can evade future precision Higgs measurements” in the large-$$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$45 limit (Meade et al., 2018). This contrast between highly testable and highly elusive realizations is a persistent feature of the subject.

A central misconception addressed explicitly in the literature is that cancellations associated with naturalness automatically favor SNR. The Twin Higgs analysis shows the opposite for pseudo-Nambu–Goldstone Higgs models with same-spin partners: the $$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$46 thermal masses cancel between $$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$47-partners, but the surviving logarithmic terms $$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$48 restore the symmetry at high temperature. For a benchmark with $$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$49 GeV and $$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$50 TeV, the quoted result is $$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$51 GeV, and the conclusion is argued to generalize to other same-spin pNGB-Higgs models (Kilic et al., 2015). A plausible implication is that SNR is not a generic consequence of ultraviolet naturalness; it depends on the detailed structure of subleading thermal terms, conserved charges, field-dependent masses, and vacuum stability constraints.

Across the current arXiv literature, EW SNR has therefore evolved from a large-$$V_{\rm eff}=V_{\rm tree}+V_{\rm CW}+V_T+V_{\rm daisy},$$52 singlet-scalar curiosity into a broader class of thermal histories encompassing scalar condensates, fermionic triggers, inert-doublet sectors, metastable broken vacua, charge-breaking intermediates, and hidden-sector constructions. What remains common to all viable realizations is the need for a quantitatively controlled finite-temperature effective potential, explicit stability and tunnelling analyses where relevant, and a model-specific account of how the negative high-temperature Higgs curvature is generated without introducing unacceptable zero-temperature pathologies (Chang et al., 2022, Biekötter et al., 2021).