We study physics at temperatures just above the QCD phase transition (Tc) using chiral (overlap) Fermions in the quenched approximation of lattice QCD. Exact zero modes of the overlap Dirac operator are localized and their frequency of occurrence drops with temperature. This is closely related to axial U (1) symmetry, which remains broken up to 2Tc. After subtracting the effects of these zero modes, chiral symmetry is restored, as indicated by the behavior of the chiral condensate ( ψψ ). The pseudoscalar and vector screening masses are close to ideal gas values.11.15.Ha, 12.38.Mh TIFR/TH/01-27, t01/076, hep-lat/0107022With new results from the Brookhaven heavy-ion collider appearing thick and fast [4], the time seems ripe for making a concerted effort to understand the dynamics of the high-temperature phase of quantum chromodynamics, namely the quark-gluon plasma. There are several puzzles that seem to have resisted a decade of efforts to understand them. The one we focus on involves the static screening of certain excitations of the plasma.It has long been understood that the screening of currents in a plasma would give us information on its excitations. Currents with certain quantum numbers excite mesons from the vacuum at low temperatures, and should exhibit deconfinement related changes above the QCD phase transition temperature (T c ) [5]. Detailed studies have shown that this indeed does happen in the vector, and axial-vector channels: the screening above T c is clearly due to nearly non-interacting quark anti-quark pairs in the medium [6,7]. On the other hand, the scalar and pseudo-scalar screening masses show more complicated behavior-strong deviations from the ideal Fermi gas, and a strong temperature dependence. This puzzling behavior is generic-it has been seen in quenched [8] and dynamical simulations with two [9] and four flavors [5][6][7]10] of staggered quarks, as well as with Wilson quarks [11]. This is the puzzle that we address and solve in this letter.The new technique we bring to bear on this problem is to use a version of lattice Fermions called overlap Fermions [12]. It has the advantage of preserving chiral symmetry on the lattice for any number of massless flavors of quarks [13]. This is in contrast to other formulations such as Wilson Fermions which break all chiral symmetries or staggered Fermions which break them partially. Since the number of pions and their nature is intimately related to the actually realized chiral symmetry on the lattice, we should expect any realization of chiral Fermions on the lattice to provide insight into the question we address.The overlap Dirac operator (D) can be defined [14] in terms of the Wilson-Dirac operator (D w ) by the relationThe computation of D −1 needs a nested series of two matrix inversions for its evaluation (each step in the numerical inversion of D involves the inversion of D † w D w ). This squaring of effort makes a study of QCD with dynamical overlap quarks very expensive. As a first step in this direction, we chose to work with quen...