Parametric models of the signal process for mX=1000GeV, mY=800GeV in their most sensitive SR. The histograms are normalized to unity. The acronym 'dof' stands for the numbers of degrees of freedom of the parametric model. The signal is modeled using a double-sided Crystal Ball (DCB) function defined as: DCB$(x)$ = \[ \begin{cases} N \cdot A_1 \cdot (B_1 - x_s)^{-m_1}, & x_s \leq -\beta_1 \\ N \cdot e^{-\frac{1}{2} x_s^2}, & -\beta_1 < x_s < \beta_2 \\ N \cdot A_2 \cdot (B_2 + x_s)^{-m_2}, & x_s \geq \beta_2 \end{cases} \] with \(x_s = \frac{x - \mu}{\sigma}\), and: \[ A_1 = \left( \frac{m_1}{\beta_1} \right)^{m_1} e^{-\frac{1}{2} \beta_1^2}, \quad B_1 = \frac{m_1}{\beta_1} - \beta_1 \] \[ A_2 = \left( \frac{m_2}{\beta_2} \right)^{m_2} e^{-\frac{1}{2} \beta_2^2}, \quad B_2 = \frac{m_2}{\beta_2} - \beta_2 \] The DCB parameters for this signal model are: \[ \begin{aligned} N &= 0.35127, & \mu &= 798.50175, & \sigma &= 7.32360 \\ \beta_1 &= 1.50996, & m_1 &= 5.30414, & \beta_2 &= 1.56253, & m_2 &= 18.3172 \end{aligned} \]

Background-only fit (red line) and signal+background fit (blue line) for the mass point hypothesis of mX=280GeV, mY=90GeV. The red line in the lower panel shows the background-only fit, which is by definition zero, and it is added as a visual aid. The most sensitive SR is shown. The choice of the background functional form is determined by the maximum likelihood fit. Category 0 is the most sensitive signal region. Drell-Yan background component is fitted as \[f_{\text{DY}}(m) = \frac{0.530965}{\sqrt{2\pi} \cdot 2.81859} \cdot \exp\left( -\frac{(m - 90.7199)^2}{2 \cdot 2.81859^2} \right)\] In the both the background-only fit and signal plus background fit, the non-resonant component of the pdf is choosing the following function. \[f(m) = \frac{1}{2.6086 \times 10^{-6}} \cdot m^{-3.82346}\] The fitted normalization factor for the function is: 23.5522 Signal fitted r value is 1.21391 in all the category simultaneous fit. The signal is fitted with a double-sided Crystal Ball (DCB) function defined as: DCB$(x)$ = \[ \begin{cases} N \cdot A_1 \cdot (B_1 - x_s)^{-m_1}, & x_s \leq -\beta_1 \\ N \cdot e^{-\frac{1}{2} x_s^2}, & -\beta_1 < x_s < \beta_2 \\ N \cdot A_2 \cdot (B_2 + x_s)^{-m_2}, & x_s \geq \beta_2 \end{cases} \] with \(x_s = \frac{x - \mu}{\sigma}\), and: \[ A_1 = \left( \frac{m_1}{\beta_1} \right)^{m_1} e^{-\frac{1}{2} \beta_1^2}, \quad B_1 = \frac{m_1}{\beta_1} - \beta_1 \] \[ A_2 = \left( \frac{m_2}{\beta_2} \right)^{m_2} e^{-\frac{1}{2} \beta_2^2}, \quad B_2 = \frac{m_2}{\beta_2} - \beta_2 \] The DCB parameters are: 2016: \[ \begin{aligned} N &= 0.6422, & \mu &= 89.9423, & \sigma &= 0.8881 \\ \beta_1 &= 1.3471, & m_1 &= 2.7908, & \beta_2 &= 1.4211, & m_2 &= 19.9999 \end{aligned} \] 2017: \[ \begin{aligned} N &= 0.8047, & \mu &= 89.9948, & \sigma &= 1.03024 \\ \beta_1 &= 1.6449, & m_1 &= 1.9425, & \beta_2 &= 1.7434, & m_2 &= 19.9982 \end{aligned} \] 2018: \[ \begin{aligned} N &= 1.1200, & \mu &= 90.0211, & \sigma &= 0.9284 \\ \beta_1 &= 1.1778, & m_1 &= 4.4273, & \beta_2 &= 1.5851, & m_2 &= 7.1198 \end{aligned} \] The plot is filled with a bin size of 5.84GeV.

Background-only fit (red line) and signal+background fit (blue line) for the mass point hypothesis of mX=280GeV, mY=90GeV. The red line in the lower panel shows the background-only fit, which is by definition zero, and it is added as a visual aid. The second most sensitive SR is shown. The choice of the background functional form is determined by the maximum likelihood fit. Category 1 is the second most sensitive signal region. Drell-Yan background component is fitted as \[f_{\text{DY}}(m) = \frac{0.532195}{\sqrt{2\pi} \cdot 3.10187} \cdot \exp\left( -\frac{(m - 89.7677)^2}{2 \cdot 3.10187^2} \right)\] In the both the background-only fit and signal plus background fit, the non-resonant component of the pdf is choosing the following function. \[f(m) = \frac{1}{14.0606} \cdot \exp\left(-\left(0.1 \cdot \frac{m}{100} + 1.73 \cdot \left(\frac{m}{100}\right)^2\right)\right)\] The fitted normalization factor for the function is: 21.6386 Signal fitted r value is 1.21391 in all the category simultaneous fit. The signal is fitted with a double-sided Crystal Ball (DCB) function defined as: DCB$(x)$ = \[ \begin{cases} N \cdot A_1 \cdot (B_1 - x_s)^{-m_1}, & x_s \leq -\beta_1 \\ N \cdot e^{-\frac{1}{2} x_s^2}, & -\beta_1 < x_s < \beta_2 \\ N \cdot A_2 \cdot (B_2 + x_s)^{-m_2}, & x_s \geq \beta_2 \end{cases} \] with \(x_s = \frac{x - \mu}{\sigma}\), and: \[ A_1 = \left( \frac{m_1}{\beta_1} \right)^{m_1} e^{-\frac{1}{2} \beta_1^2}, \quad B_1 = \frac{m_1}{\beta_1} - \beta_1 \] \[ A_2 = \left( \frac{m_2}{\beta_2} \right)^{m_2} e^{-\frac{1}{2} \beta_2^2}, \quad B_2 = \frac{m_2}{\beta_2} - \beta_2 \] The DCB parameters are: 2016: \[ \begin{aligned} N &= 0.3784, & \mu &= 89.8969, & \sigma &= 0.9189 \\ \beta_1 &= 1.1920, & m_1 &= 3.7178, & \beta_2 &= 1.3318, & m_2 &= 10.9787 \end{aligned} \] 2017: \[ \begin{aligned} N &= 0.4914, & \mu &= 89.8859, & \sigma &= 1.0728 \\ \beta_1 &= 1.3515, & m_1 &= 3.2846, & \beta_2 &= 1.4371, & m_2 &= 20.0000 \end{aligned} \] 2018: \[ \begin{aligned} N &= 0.6766, & \mu &= 89.9327, & \sigma &= 1.1213 \\ \beta_1 &= 1.9188, & m_1 &= 1.5353, & \beta_2 &= 1.9907, & m_2 &= 12.4664 \end{aligned} \] The plot is filled with a bin size of 5.84GeV.

Observed(expected) upper limits on the $\sigma\mathcal{B}$ for the X to $Y(\gamma\gamma)H(bb)$ signal with the different mass hypotheses are shown with solid(dashed) lines. The green and the yellow bands define the $\pm 68\%$ and $\pm 95\%$ uncertainty bands, respectively. For visualization purposes, the upper limit for mass points with different $m_X$ has been multiplied with a constant factor quoted on the right of each band. Mass points with $m_X$ ranging from 600 to 1000GeV are shown.

Observed(expected) upper limits on the $\sigma\mathcal{B}$ for the X 5p $Y(\gamma\gamma)H(bb)$ signal with the different mass hypotheses are shown with solid(dashed) lines. The green and the yellow bands define the $\pm 68\%$ and $\pm 95\%$ uncertainty bands, respectively. For visualization purposes, the upper limit for mass points with different $m_X$ has been multiplied with a constant factor quoted on the right of each band. Mass points with $m_X$ ranging from 240 to 550GeV are shown.

Observed(expected) upper limits on the $\sigma\mathcal{B}$ for the X to $Y(\gamma\gamma)H(bb)$ signal with the different $m_Y$ hypotheses are shown with solid(dashed) lines. The inner (green) band and the outer (yellow) band indicate the regions containing $68\%$ and $95\%$, respectively, of the distribution of limits expected under the background-only hypothesis. This plot shows the set of mass points with the lowest $m_X = 240GeV$

Observed(expected) upper limits on the $\sigma\mathcal{B}$ for the X to $Y(\gamma\gamma)H(bb)$ signal with the different $m_Y$ hypotheses are shown with solid(dashed) lines. The inner (green) band and the outer (yellow) band indicate the regions containing $68\%$ and $95\%$, respectively, of the distribution of limits expected under the background-only hypothesis. This plot shows the set of mass points with the highest $m_{X} = 1000GeV$

Expected and observed 95% CL upper limits on $\sigma(\mathrm{pp}\rightarrow\mathrm{Z}^\prime+X)\mathcal{B}(\mathrm{Z}^\prime\rightarrow e\mu)$ as a function of the Z′ mass (110–500 GeV), at $\sqrt{s} =$ 13 TeV with 138 fb$^{-1}$.

The 2D distribution of the $\Delta\Phi$ (cluster, $\mu_{trigger}$) vs the number of hits in clusters, in events from the out-of-time region.

Exclusion limits at 95% CL on the $B \rightarrow K \Phi$ as a function of the decay distance of the long-lived particle.

Exclusion limits at 95% CL on the $B \rightarrow K \Phi$ as a function of the decay distance of the long-lived particle.

Exclusion limits at 95% CL on the $B \rightarrow K \Phi$ as a function of the decay distance of the long-lived particle.

Exclusion limits at 95% CL on the $B \rightarrow K \Phi$ as a function of the decay distance of the long-lived particle.

Exclusion limits at 95% CL on the $B \rightarrow K \Phi$ as a function of the decay distance of the long-lived particle.

The estimated BR limits for a range of generator parameters: $0.3 < m_{LLP}$ < 3.0 GeV and 0 < $c\tau$ < 10000 mm.

The contour of the 2D distribution which includes the generation parameter values corresponding to the BR of 0.01.

The contour of the 2D distribution which includes the generation parameter values corresponding to the BR of 0.001.

Signal cut-flow efficiencies for the hadronic shower decay mode

Signal cut-flow efficiencies for the electromagnetic shower decay mode

The measured differences of mass differences between vector and ground states.

The measured ratios of mass differences between vector and ground states.

Predictions of a leading order (LO) QCD simulation, normalized to an integrated luminosity of 138 fb$^{-1}$. The number of events are examined within bins of the four-jet mass and the ratio $\alpha$, which is the average dijet mass divided by the four-jet mass.

The 68% probability contour in the $m_{\mathrm{4j}}$ vs. $\overline{m}_{\mathrm{2j}}$ plane from a signal simulation of a diquark with a width of 1.5% and a mass of 8.4 TeV, decaying to a pair of vector-like quarks, each with a mass of 2.1 TeV.

The 68% probability contour in the $m_{\mathrm{4j}}$ vs. $\overline{m}_{\mathrm{2j}}$ plane from a signal simulation of a diquark with a width of 5% and a mass of 8.4 TeV, decaying to a pair of vector-like quarks, each with a mass of 2.1 TeV.

The 68% probability contour in the $m_{\mathrm{4j}}$ vs. $\overline{m}_{\mathrm{2j}}$ plane from a signal simulation of a diquark with a width of 10% and a mass of 8.4 TeV, decaying to a pair of vector-like quarks, each with a mass of 2.1 TeV.

The 68% probability contour in the $m_{\mathrm{4j}}$ vs. $\alpha$ plane from a signal simulation of a diquark with a width of 1.5% and a mass of 8.4 TeV, decaying to a pair of vector-like quarks, each with a mass of 2.1 TeV.

The 68% probability contour in the $m_{\mathrm{4j}}$ vs. $\alpha$ plane from a signal simulation of a diquark with a width of 5% and a mass of 8.4 TeV, decaying to a pair of vector-like quarks, each with a mass of 2.1 TeV.

The 68% probability contour in the $m_{\mathrm{4j}}$ vs. $\alpha$ plane from a signal simulation of a diquark with a width of 10% and a mass of 8.4 TeV, decaying to a pair of vector-like quarks, each with a mass of 2.1 TeV.

Signal differential distributions as a function of four-jet mass for $\alpha_{\mathrm{true}}$ = 0.25, diquark masses of 2, 5, 8.6 TeV and various widths, for all $\alpha$ bins inclusively. The integral of each distribution has been normalized to unity.

The product of acceptance, $A$, and efficiency, $\varepsilon$, of a resonant signal with $\alpha_{\mathrm{true}}$ = 0.25 vs. the diquark mass for various diquark widths, and for all $\alpha$ bins inclusively. The acceptance is defined as the fraction of generated events passing the kinematic selection criteria, while the efficiency is the fraction of signal events satisfying $m_{\mathrm{4j}} > 1.6$ TeV. We also show the signal acceptance alone in the curves where $\varepsilon = 1$.

The four-jet mass distribution in data for 0.22 < $\alpha$ < 0.24, fitted with three background-only functions (Dijet-3p, PowExp-3p and ModDijet-3p), each with three free parameters. Examples of predicted diquark resonances with $\alpha_{\mathrm{true}}$ = 0.25, $M_{\mathrm{S}}$ = 8.6 TeV, and $\Gamma/M_{\mathrm{S}}$ = 1.5%, 10% are also shown, with cross sections equal to the observed upper limits at 95% confidence level.

The four-jet mass distribution in data for 0.24 < $\alpha$ < 0.26, fitted with three background-only functions (Dijet-3p, PowExp-3p and ModDijet-3p), each with three free parameters. Examples of predicted diquark resonances with $\alpha_{\mathrm{true}}$ = 0.25, $M_{\mathrm{S}}$ = 8.6 TeV, and $\Gamma/M_{\mathrm{S}}$ = 1.5%, 10% are also shown, with cross sections equal to the observed upper limits at 95% confidence level.

The four-jet mass distribution in data for 0.26 < $\alpha$ < 0.28, fitted with three background-only functions (Dijet-3p, PowExp-3p and ModDijet-3p), each with three free parameters. Examples of predicted diquark resonances with $\alpha_{\mathrm{true}}$ = 0.25, $M_{\mathrm{S}}$ = 8.6 TeV and $\alpha_{\mathrm{true}}$ = 0.29, $M_{\mathrm{S}}$ = 3.6 TeV, each shown for widths of $\Gamma/M_{\mathrm{S}}$ = 1.5% and 10% are also included, with cross sections equal to the observed upper limits at 95% confidence level.

The four-jet mass distribution in data for 0.28 < $\alpha$ < 0.30, fitted with three background-only functions (Dijet-3p, PowExp-3p and ModDijet-3p), each with three free parameters. Examples of predicted diquark resonances with $\alpha_{\mathrm{true}}$ = 0.25, $M_{\mathrm{S}}$ = 8.6 TeV and $\alpha_{\mathrm{true}}$ = 0.29, $M_{\mathrm{S}}$ = 3.6 TeV, each shown for widths of $\Gamma/M_{\mathrm{S}}$ = 1.5% and 10% are also included, with cross sections equal to the observed upper limits at 95% confidence level.

The four-jet mass distribution in data for 0.30 < $\alpha$ < 0.32, fitted with three background-only functions (Dijet-3p, PowExp-3p and ModDijet-3p), each with three free parameters. Examples of predicted diquark resonances with $\alpha_{\mathrm{true}}$ = 0.25, $M_{\mathrm{S}}$ = 8.6 TeV and $\alpha_{\mathrm{true}}$ = 0.29, $M_{\mathrm{S}}$ = 3.6 TeV, each shown for widths of $\Gamma/M_{\mathrm{S}}$ = 1.5% and 10% are also included, with cross sections equal to the observed upper limits at 95% confidence level.

The four-jet mass distribution in data for 0.32 < $\alpha$ < 0.34, fitted with three background-only functions (Dijet-3p, PowExp-3p and ModDijet-3p), each with three free parameters. Examples of predicted diquark resonances with $\alpha_{\mathrm{true}}$ = 0.25, $M_{\mathrm{S}}$ = 8.6 TeV, and $\Gamma/M_{\mathrm{S}}$ = 1.5%, 10% are also shown, with cross sections equal to the observed upper limits at 95% confidence level.

The inclusive four-jet mass distribution in data for $\alpha$ > 0.10, fitted with three background-only functions (Dijet-5p, PowExp-5p and ModDijet-5p), each with five free parameters. Examples of predicted diquark resonances with $\alpha_{\mathrm{true}}$ = 0.25, $M_{\mathrm{S}}$ = 8.6 TeV and $\alpha_{\mathrm{true}}$ = 0.29, $M_{\mathrm{S}}$ = 3.6 TeV, each shown for widths of $\Gamma/M_{\mathrm{S}}$ = 1.5% and 10% are also included, with cross sections equal to the observed upper limits at 95% confidence level.

The observed 95% CL upper limits on the product of the cross section, branching fraction, and acceptance for resonant production of paired dijet resonances, X, with $\alpha_{\mathrm{true}} = M_{\mathrm{X}} / M_{\mathrm{Y}} = 0.25$ and width of the initial resonance Y equal to 1.5%. The corresponding expected limits and their variations at the 1 and 2 standard deviation (s.d.) levels are also included. Limits are compared to predictions for scalar $\mathrm{S}_{\mathrm{uu}}$ and $\mathrm{S}_{\mathrm{dd}}$ diquarks with couplings to pairs of up and down quarks, $y_{\mathrm{uu}}$ and $y_{\mathrm{dd}}$, and to pairs of vector-like quarks, $y_{\chi}$ and $y_{\omega}$, set appropriately in order to generate a mediator width equal to 1.5%.

The observed 95% CL upper limits on the product of the cross section, branching fraction, and acceptance for resonant production of paired dijet resonances, X, with $\alpha_{\mathrm{true}} = M_{\mathrm{X}} / M_{\mathrm{Y}} = 0.25$ and width of the initial resonance Y equal to 5%. The corresponding expected limits and their variations at the 1 and 2 standard deviation (s.d.) levels are also included. Limits are compared to predictions for scalar $\mathrm{S}_{\mathrm{uu}}$ and $\mathrm{S}_{\mathrm{dd}}$ diquarks with couplings to pairs of up and down quarks, $y_{\mathrm{uu}}$ and $y_{\mathrm{dd}}$, and to pairs of vector-like quarks, $y_{\chi}$ and $y_{\omega}$, set appropriately in order to generate a mediator width equal to 5%.

The observed 95% CL upper limits on the product of the cross section, branching fraction, and acceptance for resonant production of paired dijet resonances, X, with $\alpha_{\mathrm{true}} = M_{\mathrm{X}} / M_{\mathrm{Y}} = 0.25$ and width of the initial resonance Y equal to 10%. The corresponding expected limits and their variations at the 1 and 2 standard deviation (s.d.) levels are also included. Limits are compared to predictions for scalar $\mathrm{S}_{\mathrm{uu}}$ and $\mathrm{S}_{\mathrm{dd}}$ diquarks with couplings to pairs of up and down quarks, $y_{\mathrm{uu}}$ and $y_{\mathrm{dd}}$, and to pairs of vector-like quarks, $y_{\chi}$ and $y_{\omega}$, set appropriately in order to generate a mediator width equal to 10%.

The observed 95% CL upper limits on the product of the cross section, branching fraction, and acceptance for resonant production of paired dijet resonances, X, with $\alpha_{\mathrm{true}} = M_{\mathrm{X}} / M_{\mathrm{Y}} = 0.11$ and width of the initial resonance Y equal to 10%. The corresponding expected limits and their variations at the 1 and 2 standard deviation (s.d.) levels are also included. Limits are compared to predictions for scalar $\mathrm{S}_{\mathrm{uu}}$ and $\mathrm{S}_{\mathrm{dd}}$ diquarks with couplings to pairs of up and down quarks, $y_{\mathrm{uu}}$ and $y_{\mathrm{dd}}$, and to pairs of vector-like quarks, $y_{\chi}$ and $y_{\omega}$, set appropriately in order to generate a mediator width equal to 10%.

The observed 95% CL upper limits on the product of the cross section, branching fraction, and acceptance for resonant production of paired dijet resonances, X, with $\alpha_{\mathrm{true}} = M_{\mathrm{X}} / M_{\mathrm{Y}} = 0.13$ and width of the initial resonance Y equal to 10%. The corresponding expected limits and their variations at the 1 and 2 standard deviation (s.d.) levels are also included. Limits are compared to predictions for scalar $\mathrm{S}_{\mathrm{uu}}$ and $\mathrm{S}_{\mathrm{dd}}$ diquarks with couplings to pairs of up and down quarks, $y_{\mathrm{uu}}$ and $y_{\mathrm{dd}}$, and to pairs of vector-like quarks, $y_{\chi}$ and $y_{\omega}$, set appropriately in order to generate a mediator width equal to 10%.

The observed 95% CL upper limits on the product of the cross section, branching fraction, and acceptance for resonant production of paired dijet resonances, X, with $\alpha_{\mathrm{true}} = M_{\mathrm{X}} / M_{\mathrm{Y}} = 0.15$ and width of the initial resonance Y equal to 10%. The corresponding expected limits and their variations at the 1 and 2 standard deviation (s.d.) levels are also included. Limits are compared to predictions for scalar $\mathrm{S}_{\mathrm{uu}}$ and $\mathrm{S}_{\mathrm{dd}}$ diquarks with couplings to pairs of up and down quarks, $y_{\mathrm{uu}}$ and $y_{\mathrm{dd}}$, and to pairs of vector-like quarks, $y_{\chi}$ and $y_{\omega}$, set appropriately in order to generate a mediator width equal to 10%.

The observed 95% CL upper limits on the product of the cross section, branching fraction, and acceptance for resonant production of paired dijet resonances, X, with $\alpha_{\mathrm{true}} = M_{\mathrm{X}} / M_{\mathrm{Y}} = 0.17$ and width of the initial resonance Y equal to 10%. The corresponding expected limits and their variations at the 1 and 2 standard deviation (s.d.) levels are also included. Limits are compared to predictions for scalar $\mathrm{S}_{\mathrm{uu}}$ and $\mathrm{S}_{\mathrm{dd}}$ diquarks with couplings to pairs of up and down quarks, $y_{\mathrm{uu}}$ and $y_{\mathrm{dd}}$, and to pairs of vector-like quarks, $y_{\chi}$ and $y_{\omega}$, set appropriately in order to generate a mediator width equal to 10%.

The observed 95% CL upper limits on the product of the cross section, branching fraction, and acceptance for resonant production of paired dijet resonances, X, with $\alpha_{\mathrm{true}} = M_{\mathrm{X}} / M_{\mathrm{Y}} = 0.19$ and width of the initial resonance Y equal to 10%. The corresponding expected limits and their variations at the 1 and 2 standard deviation (s.d.) levels are also included. Limits are compared to predictions for scalar $\mathrm{S}_{\mathrm{uu}}$ and $\mathrm{S}_{\mathrm{dd}}$ diquarks with couplings to pairs of up and down quarks, $y_{\mathrm{uu}}$ and $y_{\mathrm{dd}}$, and to pairs of vector-like quarks, $y_{\chi}$ and $y_{\omega}$, set appropriately in order to generate a mediator width equal to 10%.

The observed 95% CL upper limits on the product of the cross section, branching fraction, and acceptance for resonant production of paired dijet resonances, X, with $\alpha_{\mathrm{true}} = M_{\mathrm{X}} / M_{\mathrm{Y}} = 0.21$ and width of the initial resonance Y equal to 10%. The corresponding expected limits and their variations at the 1 and 2 standard deviation (s.d.) levels are also included. Limits are compared to predictions for scalar $\mathrm{S}_{\mathrm{uu}}$ and $\mathrm{S}_{\mathrm{dd}}$ diquarks with couplings to pairs of up and down quarks, $y_{\mathrm{uu}}$ and $y_{\mathrm{dd}}$, and to pairs of vector-like quarks, $y_{\chi}$ and $y_{\omega}$, set appropriately in order to generate a mediator width equal to 10%.

The observed 95% CL upper limits on the product of the cross section, branching fraction, and acceptance for resonant production of paired dijet resonances, X, with $\alpha_{\mathrm{true}} = M_{\mathrm{X}} / M_{\mathrm{Y}} = 0.23$ and width of the initial resonance Y equal to 10%. The corresponding expected limits and their variations at the 1 and 2 standard deviation (s.d.) levels are also included. Limits are compared to predictions for scalar $\mathrm{S}_{\mathrm{uu}}$ and $\mathrm{S}_{\mathrm{dd}}$ diquarks with couplings to pairs of up and down quarks, $y_{\mathrm{uu}}$ and $y_{\mathrm{dd}}$, and to pairs of vector-like quarks, $y_{\chi}$ and $y_{\omega}$, set appropriately in order to generate a mediator width equal to 10%.

The observed 95% CL upper limits on the product of the cross section, branching fraction, and acceptance for resonant production of paired dijet resonances, X, with $\alpha_{\mathrm{true}} = M_{\mathrm{X}} / M_{\mathrm{Y}} = 0.27$ and width of the initial resonance Y equal to 10%. The corresponding expected limits and their variations at the 1 and 2 standard deviation (s.d.) levels are also included. Limits are compared to predictions for scalar $\mathrm{S}_{\mathrm{uu}}$ and $\mathrm{S}_{\mathrm{dd}}$ diquarks with couplings to pairs of up and down quarks, $y_{\mathrm{uu}}$ and $y_{\mathrm{dd}}$, and to pairs of vector-like quarks, $y_{\chi}$ and $y_{\omega}$, set appropriately in order to generate a mediator width equal to 10%.

The observed 95% CL upper limits on the product of the cross section, branching fraction, and acceptance for resonant production of paired dijet resonances, X, with $\alpha_{\mathrm{true}} = M_{\mathrm{X}} / M_{\mathrm{Y}} = 0.29$ and width of the initial resonance Y equal to 10%. The corresponding expected limits and their variations at the 1 and 2 standard deviation (s.d.) levels are also included. Limits are compared to predictions for scalar $\mathrm{S}_{\mathrm{uu}}$ and $\mathrm{S}_{\mathrm{dd}}$ diquarks with couplings to pairs of up and down quarks, $y_{\mathrm{uu}}$ and $y_{\mathrm{dd}}$, and to pairs of vector-like quarks, $y_{\chi}$ and $y_{\omega}$, set appropriately in order to generate a mediator width equal to 10%.

The observed 95% CL upper limits on the product of the cross section, branching fraction, and acceptance for resonant production of paired dijet resonances, X, with $\alpha_{\mathrm{true}} = M_{\mathrm{X}} / M_{\mathrm{Y}} = 0.31$ and width of the initial resonance Y equal to 10%. The corresponding expected limits and their variations at the 1 and 2 standard deviation (s.d.) levels are also included. Limits are compared to predictions for scalar $\mathrm{S}_{\mathrm{uu}}$ and $\mathrm{S}_{\mathrm{dd}}$ diquarks with couplings to pairs of up and down quarks, $y_{\mathrm{uu}}$ and $y_{\mathrm{dd}}$, and to pairs of vector-like quarks, $y_{\chi}$ and $y_{\omega}$, set appropriately in order to generate a mediator width equal to 10%.

The observed 95% CL upper limits on the product of the cross section, branching fraction, and acceptance for resonant production of paired dijet resonances, X, with $\alpha_{\mathrm{true}} = M_{\mathrm{X}} / M_{\mathrm{Y}} = 0.33$ and width of the initial resonance Y equal to 10%. The corresponding expected limits and their variations at the 1 and 2 standard deviation (s.d.) levels are also included. Limits are compared to predictions for scalar $\mathrm{S}_{\mathrm{uu}}$ and $\mathrm{S}_{\mathrm{dd}}$ diquarks with couplings to pairs of up and down quarks, $y_{\mathrm{uu}}$ and $y_{\mathrm{dd}}$, and to pairs of vector-like quarks, $y_{\chi}$ and $y_{\omega}$, set appropriately in order to generate a mediator width equal to 10%.

The observed 95% CL upper limits on the product of the cross section, branching fraction, and acceptance for resonant production of paired dijet resonances, X, with $\alpha_{\mathrm{true}} = M_{\mathrm{X}} / M_{\mathrm{Y}} = 0.42$ and width of the initial resonance Y equal to 10%. The corresponding expected limits and their variations at the 1 and 2 standard deviation (s.d.) levels are also included. Limits are compared to predictions for scalar $\mathrm{S}_{\mathrm{uu}}$ and $\mathrm{S}_{\mathrm{dd}}$ diquarks with couplings to pairs of up and down quarks, $y_{\mathrm{uu}}$ and $y_{\mathrm{dd}}$, and to pairs of vector-like quarks, $y_{\chi}$ and $y_{\omega}$, set appropriately in order to generate a mediator width equal to 10%.

The observed 95% CL upper limits on the product of the cross section, branching fraction, and acceptance for resonant production of paired dijet resonances, X, with $\alpha_{\mathrm{true}} = M_{\mathrm{X}} / M_{\mathrm{Y}} = 0.11$ and widths of the initial resonance Y equal to 1.5, 5, and 10%. Limits are compared to predictions for scalar $\mathrm{S}_{\mathrm{uu}}$ and $\mathrm{S}_{\mathrm{dd}}$ diquarks with couplings to pairs of up and down quarks, $y_{\mathrm{uu}}$ and $y_{\mathrm{dd}}$, and to pairs of vector-like quarks, $y_{\chi}$ and $y_{\omega}$, set appropriately in order to generate the corresponding widths.

The observed 95% CL upper limits on the product of the cross section, branching fraction, and acceptance for resonant production of paired dijet resonances, X, with $\alpha_{\mathrm{true}} = M_{\mathrm{X}} / M_{\mathrm{Y}} = 0.13$ and widths of the initial resonance Y equal to 1.5, 5, and 10%. Limits are compared to predictions for scalar $\mathrm{S}_{\mathrm{uu}}$ and $\mathrm{S}_{\mathrm{dd}}$ diquarks with couplings to pairs of up and down quarks, $y_{\mathrm{uu}}$ and $y_{\mathrm{dd}}$, and to pairs of vector-like quarks, $y_{\chi}$ and $y_{\omega}$, set appropriately in order to generate the corresponding widths.

The observed 95% CL upper limits on the product of the cross section, branching fraction, and acceptance for resonant production of paired dijet resonances, X, with $\alpha_{\mathrm{true}} = M_{\mathrm{X}} / M_{\mathrm{Y}} = 0.15$ and widths of the initial resonance Y equal to 1.5, 5, and 10%. Limits are compared to predictions for scalar $\mathrm{S}_{\mathrm{uu}}$ and $\mathrm{S}_{\mathrm{dd}}$ diquarks with couplings to pairs of up and down quarks, $y_{\mathrm{uu}}$ and $y_{\mathrm{dd}}$, and to pairs of vector-like quarks, $y_{\chi}$ and $y_{\omega}$, set appropriately in order to generate the corresponding widths.

The observed 95% CL upper limits on the product of the cross section, branching fraction, and acceptance for resonant production of paired dijet resonances, X, with $\alpha_{\mathrm{true}} = M_{\mathrm{X}} / M_{\mathrm{Y}} = 0.17$ and widths of the initial resonance Y equal to 1.5, 5, and 10%. Limits are compared to predictions for scalar $\mathrm{S}_{\mathrm{uu}}$ and $\mathrm{S}_{\mathrm{dd}}$ diquarks with couplings to pairs of up and down quarks, $y_{\mathrm{uu}}$ and $y_{\mathrm{dd}}$, and to pairs of vector-like quarks, $y_{\chi}$ and $y_{\omega}$, set appropriately in order to generate the corresponding widths.

The observed 95% CL upper limits on the product of the cross section, branching fraction, and acceptance for resonant production of paired dijet resonances, X, with $\alpha_{\mathrm{true}} = M_{\mathrm{X}} / M_{\mathrm{Y}} = 0.19$ and widths of the initial resonance Y equal to 1.5, 5, and 10%. Limits are compared to predictions for scalar $\mathrm{S}_{\mathrm{uu}}$ and $\mathrm{S}_{\mathrm{dd}}$ diquarks with couplings to pairs of up and down quarks, $y_{\mathrm{uu}}$ and $y_{\mathrm{dd}}$, and to pairs of vector-like quarks, $y_{\chi}$ and $y_{\omega}$, set appropriately in order to generate the corresponding widths.

The observed 95% CL upper limits on the product of the cross section, branching fraction, and acceptance for resonant production of paired dijet resonances, X, with $\alpha_{\mathrm{true}} = M_{\mathrm{X}} / M_{\mathrm{Y}} = 0.21$ and widths of the initial resonance Y equal to 1.5, 5, and 10%. Limits are compared to predictions for scalar $\mathrm{S}_{\mathrm{uu}}$ and $\mathrm{S}_{\mathrm{dd}}$ diquarks with couplings to pairs of up and down quarks, $y_{\mathrm{uu}}$ and $y_{\mathrm{dd}}$, and to pairs of vector-like quarks, $y_{\chi}$ and $y_{\omega}$, set appropriately in order to generate the corresponding widths.

The observed 95% CL upper limits on the product of the cross section, branching fraction, and acceptance for resonant production of paired dijet resonances, X, with $\alpha_{\mathrm{true}} = M_{\mathrm{X}} / M_{\mathrm{Y}} = 0.23$ and widths of the initial resonance Y equal to 1.5, 5, and 10%. Limits are compared to predictions for scalar $\mathrm{S}_{\mathrm{uu}}$ and $\mathrm{S}_{\mathrm{dd}}$ diquarks with couplings to pairs of up and down quarks, $y_{\mathrm{uu}}$ and $y_{\mathrm{dd}}$, and to pairs of vector-like quarks, $y_{\chi}$ and $y_{\omega}$, set appropriately in order to generate the corresponding widths.

The observed 95% CL upper limits on the product of the cross section, branching fraction, and acceptance for resonant production of paired dijet resonances, X, with $\alpha_{\mathrm{true}} = M_{\mathrm{X}} / M_{\mathrm{Y}} = 0.27$ and widths of the initial resonance Y equal to 1.5, 5, and 10%. Limits are compared to predictions for scalar $\mathrm{S}_{\mathrm{uu}}$ and $\mathrm{S}_{\mathrm{dd}}$ diquarks with couplings to pairs of up and down quarks, $y_{\mathrm{uu}}$ and $y_{\mathrm{dd}}$, and to pairs of vector-like quarks, $y_{\chi}$ and $y_{\omega}$, set appropriately in order to generate the corresponding widths.

The observed 95% CL upper limits on the product of the cross section, branching fraction, and acceptance for resonant production of paired dijet resonances, X, with $\alpha_{\mathrm{true}} = M_{\mathrm{X}} / M_{\mathrm{Y}} = 0.29$ and widths of the initial resonance Y equal to 1.5, 5, and 10%. Limits are compared to predictions for scalar $\mathrm{S}_{\mathrm{uu}}$ and $\mathrm{S}_{\mathrm{dd}}$ diquarks with couplings to pairs of up and down quarks, $y_{\mathrm{uu}}$ and $y_{\mathrm{dd}}$, and to pairs of vector-like quarks, $y_{\chi}$ and $y_{\omega}$, set appropriately in order to generate the corresponding widths.

The observed 95% CL upper limits on the product of the cross section, branching fraction, and acceptance for resonant production of paired dijet resonances, X, with $\alpha_{\mathrm{true}} = M_{\mathrm{X}} / M_{\mathrm{Y}} = 0.31$ and widths of the initial resonance Y equal to 1.5, 5, and 10%. Limits are compared to predictions for scalar $\mathrm{S}_{\mathrm{uu}}$ and $\mathrm{S}_{\mathrm{dd}}$ diquarks with couplings to pairs of up and down quarks, $y_{\mathrm{uu}}$ and $y_{\mathrm{dd}}$, and to pairs of vector-like quarks, $y_{\chi}$ and $y_{\omega}$, set appropriately in order to generate the corresponding widths.

The observed 95% CL upper limits on the product of the cross section, branching fraction, and acceptance for resonant production of paired dijet resonances, X, with $\alpha_{\mathrm{true}} = M_{\mathrm{X}} / M_{\mathrm{Y}} = 0.33$ and widths of the initial resonance Y equal to 1.5, 5, and 10%. Limits are compared to predictions for scalar $\mathrm{S}_{\mathrm{uu}}$ and $\mathrm{S}_{\mathrm{dd}}$ diquarks with couplings to pairs of up and down quarks, $y_{\mathrm{uu}}$ and $y_{\mathrm{dd}}$, and to pairs of vector-like quarks, $y_{\chi}$ and $y_{\omega}$, set appropriately in order to generate the corresponding widths.

The observed 95% CL upper limits on the product of the cross section, branching fraction, and acceptance for resonant production of paired dijet resonances, X, with $\alpha_{\mathrm{true}} = M_{\mathrm{X}} / M_{\mathrm{Y}} = 0.42$ and widths of the initial resonance Y equal to 1.5, 5, and 10%. Limits are compared to predictions for scalar $\mathrm{S}_{\mathrm{uu}}$ and $\mathrm{S}_{\mathrm{dd}}$ diquarks with couplings to pairs of up and down quarks, $y_{\mathrm{uu}}$ and $y_{\mathrm{dd}}$, and to pairs of vector-like quarks, $y_{\chi}$ and $y_{\omega}$, set appropriately in order to generate the corresponding widths.

Observed local $p$-value for a four-jet resonance, Y, decaying to a pair of dijet resonances, X, with $\alpha_{\mathrm{true}} = M_{\mathrm{X}}/M_{\mathrm{Y}}$ = 0.25, and various widths of Y superimposed.

Observed local $p$-value for a four-jet resonance, Y, decaying to a pair of dijet resonances, X, with $\alpha_{\mathrm{true}} = M_{\mathrm{X}}/M_{\mathrm{Y}}$ = 0.29, and various widths of Y superimposed.

The table presents the cumulative cutflow for a signal with $M_{\mathrm{S}} = 2.0$ TeV, $M_{\chi} = 0.5$ TeV and different width hypotheses ($\Gamma/M_{\mathrm{S}} =$ 1.5, 5, and 10%). Each row corresponds to the number of signal events that survive all cuts up to and including the one listed. The percentage in parentheses shows the efficiency of the current cut alone, defined as the ratio of the number of events that survive all cuts up to and including this one to the number of events that survived all previous cuts.

The table presents the cumulative cutflow for a signal with $M_{\mathrm{S}} = 5.0$ TeV, $M_{\chi} = 1.25$ TeV and different width hypotheses ($\Gamma/M_{\mathrm{S}} =$ 1.5, 5, and 10%). Each row corresponds to the number of signal events that survive all cuts up to and including the one listed. The percentage in parentheses shows the efficiency of the current cut alone, defined as the ratio of the number of events that survive all cuts up to and including this one to the number of events that survived all previous cuts.

The table presents the cumulative cutflow for a signal with $M_{\mathrm{S}} = 8.6$ TeV, $M_{\chi} = 2.15$ TeV and different width hypotheses ($\Gamma/M_{\mathrm{S}} =$ 1.5, 5, and 10%). Each row corresponds to the number of signal events that survive all cuts up to and including the one listed. The percentage in parentheses shows the efficiency of the current cut alone, defined as the ratio of the number of events that survive all cuts up to and including this one to the number of events that survived all previous cuts.

Observed model-independent upper limits at 90 % CL of $\mathcal{B}(K^{+}\rightarrow\pi^{+}X)$ as function of $X$ mass, for several $X$ lifetime hypotheses, assuming $X$ decays to visible SM particles.

Di-muon mass spectrum of selected $K^{+}\rightarrow\pi^{+}\mu^{+}\mu^{-}$ events from 2017–2018 data.

Expected and observed model-independent upper limits at 90 % CL for $\mathcal{B}(K^{+}\rightarrow\pi^{+}X)\times\mathcal{B}(X\rightarrow\mu^{+}\mu^{-})$ as a function of $m_{X}$ for $\tau_{X}=0$.

Observed model-independent upper limits at 90 % CL for $\mathcal{B}(K^{+}\rightarrow\pi^{+}X)\times\mathcal{B}(X\rightarrow\mu^{+}\mu^{-})$ for several $\tau_{X}$ values.

Branching ratio of the $K^{+}\rightarrow\pi^{+}A^{\prime}$ decay divided by the kinetic mixing coupling squared, $\varepsilon^{2}$, as a function of $m_{A^{\prime}}$ for the BC2 model with dark photon $A^{\prime}$.

Excluded region, at 90 % CL, of the parameter space $(m_{A^{\prime}},\varepsilon)$ for a dark photon $A^{\prime}$, decaying invisibly, in the BC2 model for the NA62 $K^{+}\rightarrow\pi^{+}X_{\rm inv}$ search.

Excluded region, at 90 % CL, of the parameter space $(m_{A^{\prime}},\varepsilon)$ for a dark photon $A^{\prime}$, decaying invisibly, in the BC2 model for the NA62 $\pi^{0}\rightarrow{\rm inv}$ search.

Branching ratio of the $K^{+}\rightarrow\pi^{+}S$ decay divided by $\text{sin}^{2}\theta$, as a function of $m_{S}$.

Excluded regions, at 90 % CL, of the parameter space $(m_{S},\rm{sin}^{2}\theta)$ for a dark scalar $S$, in the BC4-inv model for the NA62 $K^{+}\rightarrow\pi^{+}X_{\rm inv}$ search.

Excluded regions, at 90 % CL, of the parameter space $(m_{S},\rm{sin}^{2}\theta)$ for a dark scalar $S$, in the BC4-inv model for the NA62 $\pi^{0}\rightarrow{\rm inv}$ search.

Excluded regions, at 90 % CL, of the parameter space $(m_{S},\rm{sin}^{2}\theta$) for a dar scalar S, in the BC4 model, from the NA62 $K^{+}\rightarrow\pi^{+}X_{\rm inv}$ search.

Excluded regions, at 90 % CL, of the parameter space $(m_{S},\rm{sin}^{2}\theta$) for a dar scalar S, in the BC4 model, from the NA62 $K^{+}\rightarrow\pi^{+}\mu^{+}\mu^{-}$ study.

Branching ratio of the $K^{+}\rightarrow\pi^{+}a$ decay divided by $(C_{ff}/\Lambda)^{2}$, as a function of $m_{a}$, assuming Λ = 1 TeV.

Excluded regions, at 90 % CL, of the parameter space $(m_{a}, C_{ff}/\Lambda)$ for an ALP $a$, in the BC10-inv model, evaluated assuming $\Lambda = 1$ TeV, for the NA62 $K^{+}\rightarrow\pi^{+}X_{\rm inv}$ search.

Excluded regions, at 90 % CL, of the parameter space $(m_{a}, C_{ff}/\Lambda)$ for an ALP $a$, in the BC10-inv model, evaluated assuming $\Lambda = 1$ TeV, for the NA62 $\pi^{0}\rightarrow\rm{inv}$ search.

Excluded regions, at 90 % CL, of the parameter space $(m_{a}, C_{ff}/\lambda)$ for an ALP $a$, in the BC10 model, evaluated assuming $\Lambda = 1$ TeV for the NA62 $K^{+}\rightarrow\pi^{+}X_{\rm inv}$ search.

Excluded regions, at 90 % CL, of the parameter space $(m_{a}, C_{ff}/\lambda)$ for an ALP $a$, in the BC10 model, evaluated assuming $\Lambda = 1$ TeV for the NA62 $\pi^{0}\rightarrow\rm{inv}$ search.

Excluded regions, at 90 % CL, of the parameter space $(m_{a}, C_{ff}/\lambda)$ for an ALP $a$, in the BC10 model, evaluated assuming $\Lambda = 1$ TeV for the NA62 $K^{+}\rightarrow\pi^{+}\mu^{+}\mu^{-}$ study.

Branching ratio of the $K^{+}\rightarrow\pi^{+}a$ decay divided by $(C_{GG}/\Lambda)^{2}$, as a function of $m_{a}$, assuming $\Lambda = 1$ TeV.

Excluded regions, at 90 % CL, of the parameter space $(m_{a}, C_{GG}/\Lambda)$ for an ALP $a$, of the BC11 model, evaluated assuming $\Lambda = 1$ TeV, from the NA62 search for $K^{+}\rightarrow\pi^{+}X_{\rm inv}$.

Excluded regions, at 90 % CL, of the parameter space $(m_{a}, C_{GG}/\Lambda)$ for an ALP $a$, of the BC11 model, evaluated assuming $\Lambda = 1$ TeV, from the NA62 search for $\pi^{0}\rightarrow\rm{inv}$.

Excluded regions, at 90 % CL, of the parameter space $(m_{a}, C_{GG}/\Lambda)$ for an ALP $a$, of the BC11 model, evaluated assuming $\Lambda = 1$ TeV, from the NA62 $K^{+}\rightarrow\pi^{+}\gamma\gamma$ study.

$\xi$ distributions for different jet $p_T$ bins.

$\xi$ distributions for different jet $p_T$ bins.

The $j_T/p_{T}^{\rm jet}$ distributions for different jet $p_T$ bins.

The $j_T/p_{T}^{\rm jet}$ distributions for different jet $p_T$ bins.

$r$ distributions for different jet $p_T$ bins.

Distributions of discriminant for the $2_{m}^{+}$ and $0^{-}$ models.