Computations of Galois representations associated to modular forms of level one (Q2875430)

``Galois representations over finite fields attached to modular forms of level one can, in almost all cases, be computed in polynomial time in the weight of the form and in the size of the finite field.'' (This was proved in the survey book by \textit{B. Edixhoven} (ed.) and \textit{J.-M. Couveignes} (ed.) [Computational aspects of modular forms and Galois representations. How one can compute in polynomial time the value of Ramanujan's tau at a prime. Princeton, NJ: Princeton University Press (2011; Zbl 1216.11004)]). Hence, the coefficients of the \(q\)-expansions of such modular forms can be computed fast via congruences. In his thesis \textit{J. Bosman} developed the computational methods for this approach in order to determine the coefficients \(\tau(p)\) in the \(q\)-expansion of Ramanujan's tau function for primes \(p\) [Explicit computations with modular Galois representations. Leiden: University of Leiden (PhD Thesis) (2008)]. Especially, he computed polynomials for projective representations of level one forms of weight \(k\leq 22\) and primes \(\ell \leq 23\).NEWLINENEWLINENEWLINEIn this paper, the author improves that method in case \(\gcd(k-2,\ell+1)>2\). Whereas Bosman used the modular curve \(X_1(\ell)\) the author can work with a suitable modular curve \(\Gamma\) of smaller genus. In this way the needed precision of the calculations becomes smaller. In Sections 2 to 4, the algorithm is described, Galois theory of modular curves is discussed, and the algorithm using \(\Gamma\) rather than \(X_1(\ell)\) is explained, i.e., essentially the determining of \(\Gamma\). In Section 5, the polynomials for the new examples with { small \((k,\ell) \in \{ (12,31), (16,29), (20,31), (22,31)\}\)} (of degree \(\ell +1\) each) are presented. For the calculations a precision of about 4000 bits was needed and the computation time was about 10 days in each case.NEWLINENEWLINEAlso, the values modulo 31 of Ramanujan's tau function were calculated for several large primes resulting in \(\tau(n) \neq 0\) \quad for all \(n< 982149821766199295999\) improving a former result of Bosman by a factor of roughly 43.