arXiv2026
We study the sensitivity of a future $μp$ collider to singly produced singlet vector-like $T$ quarks in the $T\to Wb$ decay mode across both fully hadronic ($bjj$) and leptonic ($b\ellν$) final states at $\sqrt{s}=5.29$, $6.48$, and $9.16~\mathrm{TeV}$. Utilizing resolved-object reconstruction and TMVA-based multivariate classifiers (BDT, BDTG, MLP, and Likelihood), we perform a fast-simulation scan over $m_T=2$--$5~\mathrm{TeV}$ and the $(g^{*},m_T)$ parameter space. To quantify improvements, we compare this multivariate selection against the closest cut-based baseline at a benchmark point of $m_T=3~\mathrm{TeV}$, $\sqrt{s}=9.16~\mathrm{TeV}$, and $\mathcal{L}=3000~\mathrm{fb}^{-1}$, incorporating a $20\%$ background-normalization uncertainty via the Asimov significance $Z_A$. At this benchmark, the hadronic channel exhibits a substantial enhancement in both $S/\sqrt{S+B}$ and $Z_A$, improving from $18.05$ and $1.44$ to $273.51$ and $42.24$, respectively. Similarly, the leptonic channel improves from $3.51$ and $0.330$ to $113.85$ and $21.51$. In the full parameter scan, the $\sqrt{s}=9.16~\mathrm{TeV}$ configuration provides the strongest sensitivity at $\mathcal{L}=3000~\mathrm{fb}^{-1}$. The hadronic channel excludes (discovers) regions with $g^{*}\in[0.20,0.50]$ ($[0.30,0.50]$) up to $m_T\approx4.0~(3.5)~\mathrm{TeV}$, while the leptonic channel remains sensitive up to $m_T\approx5.0~\mathrm{TeV}$ for $g^{*}\in[0.10,0.50]$. These gains arise from the higher-purity multivariate signal regions evaluated under a common significance prescription, and not from a change of the collider configuration. The projections assume resolved-object reconstruction and fast detector simulation, and the absolute reach should be read with these assumptions in mind.