#include using namespace std; #define rep(i,n) for(ll i=0;i=0;i--) #define perl(i,r,l) for(ll i=r-1;i>=l;i--) #define fi first #define se second #define pb push_back #define ins insert #define pqueue(x) priority_queue,greater> #define all(x) (x).begin(),(x).end() #define CST(x) cout<; using vvl=vector>; using pl=pair; using vpl=vector; using vvpl=vector; const ll MOD=1000000007; const ll MOD9=998244353; const int inf=1e9+10; const ll INF=4e18; //const ll dy[8]={-1,0,1,0,1,1,-1,-1}; //const ll dx[8]={0,-1,0,1,1,-1,1,-1}; template inline bool chmin(T& a, T b) { if (a > b) { a = b; return true; } return false; } template inline bool chmax(T& a, T b) { if (a < b) { a = b; return true; } return false; } namespace NTT { // int32型のmodが取れるFFT。auto c=NTT::mul(a,b,mod)で受け取り。TIME指定。 // ChineseRemと組み合わせてlong longにもできる std::vector tmp; size_t sz = 1; inline int powMod(int n, int p, int m) { int res = 1; while (p) { if (p & 1) res = (ll)res * n % m; n = (ll)n * n % m; p >>= 1; } return (int)res; } inline int invMod(int n, int m) { return powMod(n, m - 2, m); } template struct NTTPart { static std::vector ntt(std::vector a, bool inv = false) { size_t mask = sz - 1; size_t p = 0; for (size_t i = sz >> 1; i >= 1; i >>= 1) { auto& cur = (p & 1) ? tmp : a; auto& nex = (p & 1) ? a : tmp; int e = powMod(PrimitiveRoot, (Mod - 1) / sz * i, Mod); if (inv) e = invMod(e, Mod); int w = 1; for (size_t j = 0; j < sz; j += i) { for (size_t k = 0; k < i; ++k) { nex[j + k] = (cur[((j << 1) & mask) + k] + (ll)w * cur[(((j << 1) + i) & mask) + k]) % Mod; } w = (ll)w * e % Mod; } ++p; } if (p & 1) std::swap(a, tmp); if (inv) { int invSz = invMod(sz, Mod); for (size_t i = 0; i < sz; ++i) a[i] = (ll)a[i] * invSz % Mod; } return a; } static std::vector mul(std::vector a, std::vector b) { a = ntt(a); b = ntt(b); for (size_t i = 0; i < sz; ++i) a[i] = (ll)a[i] * b[i] % Mod; a = ntt(a, true); return a; } }; constexpr int M[] = {1224736769, 469762049, 167772161}; constexpr int PR[] = {3, 3, 3}; constexpr int NTT_CONVOLUTION_TIME = 3; /* X := max(a)*max(b)*max(|a|, |b|) のとき, NTT_CONVOLUTION_TIME <- 1: X < 1224736769 = 1.2*10^ 9 ~ 2^30 NTT_CONVOLUTION_TIME <- 2: X < 575334854091079681 = 5.8*10^17 ~ 2^59 NTT_CONVOLUTION_TIME <- 3: X < 2^86 (32bit * 32bit * 10^7くらいまで) */ inline void garner(std::vector *c, int mod) { if (NTT_CONVOLUTION_TIME == 1) { for(auto& x : c[0]) x %= mod; } else if (NTT_CONVOLUTION_TIME == 2) { const int r01 = invMod(M[0], M[1]); for (size_t i = 0; i < sz; ++i) { c[1][i] = (c[1][i] - c[0][i]) * (ll)r01 % M[1]; if (c[1][i] < 0) c[1][i] += M[1]; c[0][i] = (c[0][i] + (ll)c[1][i] * M[0]) % mod; } } else if (NTT_CONVOLUTION_TIME == 3) { const int R01 = invMod(M[0], M[1]); const int R02 = invMod(M[0], M[2]); const int R12 = invMod(M[1], M[2]); const int M01 = (ll)M[0] * M[1] % mod; for (size_t i = 0; i < sz; ++i) { c[1][i] = (c[1][i] - c[0][i]) * (ll)R01 % M[1]; if (c[1][i] < 0) c[1][i] += M[1]; c[2][i] = ((c[2][i] - c[0][i]) * (ll)R02 % M[2] - c[1][i]) * R12 % M[2]; if (c[2][i] < 0) c[2][i] += M[2]; c[0][i] = (c[0][i] + (ll)c[1][i] * M[0] + (ll)c[2][i] * M01) % mod; } } } std::vector mul(std::vector a, std::vector b, int mod) { for (auto& x : a) x %= mod; for (auto& x : b) x %= mod; size_t m = a.size() + b.size() - 1; sz = 1; while (m > sz) sz <<= 1; tmp.resize(sz); a.resize(sz, 0); b.resize(sz, 0); std::vector c[NTT_CONVOLUTION_TIME]; if (NTT_CONVOLUTION_TIME >= 1) c[0] = NTTPart::mul(a, b); if (NTT_CONVOLUTION_TIME >= 2) c[1] = NTTPart::mul(a, b); if (NTT_CONVOLUTION_TIME >= 3) c[2] = NTTPart::mul(a, b); for (auto& v : c) v.resize(m); garner(c, mod); return c[0]; } }; // !!! CHECK NTT_CONVOLUTION_TIME !!! int main(){ ll n,q;cin >> n >>q; vector v={1}; rep(i,n){ ll a;cin >> a;a--;a%=MOD9; int k=a; vector p={1,k}; v=NTT::mul(v,p,MOD9); //for(auto z:v)cout << z <<" ";cout << endl; } while(q--){ ll x;cin >> x; cout << v[n-x] <