Local Gan–Gross–Prasad conjecture for tempered representations of real unitary groups

At least 10 years old · documented by

Let ϕn∈Φtemp(Un(R))\phi_n \in \Phi_\mathrm{temp}(\mathrm{U}_{n}(\mathbb{R})) and ϕn+1∈Φtemp(Un+1(R))\phi_{n+1} \in \Phi_\mathrm{temp}(\mathrm{U}_{n+1}(\mathbb{R})) have decompositions

ϕn=(m1χ2α1⊕⋯⊕muχ2αu)⊕(ξ1⊕⋯⊕ξv)⊕(cξ1−1⊕⋯⊕cξv−1),\phi_{n}=(m_1\chi_{2\alpha_1}\oplus\dots\oplus m_u\chi_{2\alpha_u})\oplus(\xi_1\oplus\dots\oplus\xi_v)\oplus({}^c\xi_1^{-1}\oplus\dots\oplus{}^c\xi_v^{-1}), ϕn+1=(m1′χ2β1⊕⋯⊕mu′′χ2βu′)⊕(ξ1′⊕⋯⊕ξv′′)⊕(cξ1′−1⊕⋯⊕cξv′′−1),\phi_{n+1}=(m'_1\chi_{2\beta_1}\oplus\dots\oplus m'_{u'}\chi_{2\beta_{u'}})\oplus(\xi'_1\oplus\dots\oplus\xi'_{v'})\oplus({}^c{\xi'_1}^{-1}\oplus\dots\oplus{}^c{\xi'_{v'}}^{-1}),

where αi,βj∈12Z\alpha_i,\beta_j\in\frac{1}{2}\mathbb{Z} satisfy 2αi≡n−1(mod2)2\alpha_i\equiv n-1\pmod 2 and 2βj≡n(mod2)2\beta_j\equiv n\pmod 2, the mim_i and mj′m'_j are the corresponding positive multiplicities, m1+⋯+mu+2v=nm_1+\dots+m_u+2v=n, m1′+⋯+mu′′+2v′=n+1m'_1+\dots+m'_{u'}+2v'=n+1, and the ξi\xi_i and ξj′\xi'_j are unitary characters of C×\mathbb{C}^{\times} not of the indicated character forms. Local Gan–Gross–Prasad conjecture. There exists a unique pair (πn,πn+1)∈Πϕn×Πϕn+1(\pi_n,\pi_{n+1})\in\Pi_{\phi_n}\times\Pi_{\phi_{n+1}} such that (πn,πn+1)(\pi_n,\pi_{n+1}) is a pair of representations of a relevant pair (Un(R),Un+1(R))(\mathrm{U}_n(\mathbb{R}),\mathrm{U}_{n+1}(\mathbb{R})) and

Hom⁡ΔUn(R)(πn⊗πn+1,C)≠0.\operatorname{Hom}_{\Delta\mathrm{U}_{n}(\mathbb{R})}(\pi_n\otimes\pi_{n+1},\mathbb{C})\ne0.

Moreover, for e2αi∈Aϕne_{2\alpha_i}\in A_{\phi_n} and e2βj∈Aϕn+1e_{2\beta_j}\in A_{\phi_{n+1}},

J(πn)(e2αi)=ε(χ2αi⊗ϕn+1,ψ−2C),J(\pi_n)(e_{2\alpha_i})=\varepsilon(\chi_{2\alpha_i}\otimes\phi_{n+1},\psi_{-2}^{\mathbb{C}}), J(πn+1)(e2βj)=ε(ϕn⊗χ2βj,ψ−2C).J(\pi_{n+1})(e_{2\beta_j})=\varepsilon(\phi_n\otimes\chi_{2\beta_j},\psi_{-2}^{\mathbb{C}}).

Here relevance means (U(p,q),U(p+1,q))(\mathrm{U}(p,q),\mathrm{U}(p+1,q)) when nn is even and (U(p,q),U(p,q+1))(\mathrm{U}(p,q),\mathrm{U}(p,q+1)) when nn is odd. This conjecture predicts exactly when the restriction multiplicity is nonzero; Sun–Zhu proved that the multiplicity is at most one, while the asserted existence, uniqueness, and epsilon-factor characterization are the remaining content.

References

Primary source

Hiraku Atobe, “On the non-vanishing of theta liftings of tempered representations of U(p,q)”, arXiv:1610.07794 (2016).

Additional references

2 papers in this index state this conjecture (2015–2016). The statement above is taken from the most recent of them; the others are arXiv:1502.03528.

Progress summary

Never refreshed

Nothing recorded yet. Refresh searches the literature and the public web for attempts on this problem, and writes the first summary here.

Solutions 0

No solutions have been posted yet.