Fig. 1. (a) Experimental setup. Three coupling beams, E2, E2′, and E2′′, from the same laser source intersect inside the vapor cell to establish a hexagonal interference pattern acting as the coupling field Ec. The angle between any two of the coupling beams is 2 θ≈0.4  deg. The focal length of the imaging lens is 200 mm, and the distance between the lens and the CCD camera is 400 mm. (b) The energy-level configuration driven by the probe and coupling fields; (c) projection of the four beams (before entering the cell) on the x−y plane; calculated susceptibility at (d), (e) Δ1=−60  MHz and (f), (g) Δ1=−110  MHz with Δ2=−100  MHz, respectively.

Fig. 2. Numerical simulations of a complex honeycomb photonic lattice. (a) Dispersion in the Hermitian case showing the lowest bands (s and mixed p/d band); (b) dispersion in the non-Hermitian case with the maximal intensity coming from the lowest-decay state at the Γ point of the p/d band; (c), (d) spatial profiles of the Γ states of the s and p/d bands, responsible for the lattice switching. Here, a and a* are the direct and reciprocal lattice constants.

Fig. 3. Output probe patterns at different probe detunings. (a) Experimentally established coupling field; (b) observed discretized probe beam at different Δ1 with a fixed Δ2=−100  MHz. The corresponding simulations are shown in (c).

Fig. 4. Observed self-imaging effect of the output probe beam at different probe detunings. (a) Δ1=−90  MHz and (b) Δ1=−110  MHz.

Fig. 5. PT-symmetric-like transition measured by the symmetry factor as a function of the relative non-Hermiticity controlled by the detuning; black solid line, square root scaling; black dots, numerical simulations; red dots, experiment (error bars indicate the uncertainty).