#include <bits/stdc++.h>
using ll = long long;
using uint = unsigned int;
using ull = unsigned long long;
using ld = long double;
template<typename T> using max_heap = std::priority_queue<T>;
template<typename T> using min_heap = std::priority_queue<T, std::vector<T>, std::greater<T>>;
constexpr int popcount(const ull v) { return v ? __builtin_popcountll(v) : 0; }
constexpr int log2p1(const ull v) { return v ? 64 - __builtin_clzll(v) : 0; }
constexpr int lsbp1(const ull v) { return __builtin_ffsll(v); }
constexpr int clog(const ull v) { return v ? log2p1(v - 1) : 0; }
constexpr ull ceil2(const ull v) { return 1ULL << clog(v); }
constexpr ull floor2(const ull v) { return v ? (1ULL << (log2p1(v) - 1)) : 0ULL; }
constexpr bool btest(const ull mask, const int ind) { return (mask >> ind) & 1ULL; }
template<typename T> bool chmin(T& a, const T& b) { return (a > b ? a = b, true : false); }
template<typename T> bool chmax(T& a, const T& b) { return (a < b ? a = b, true : false); }
template<typename T> constexpr T inf_v = std::numeric_limits<T>::max() / 4;
template<typename Real> constexpr Real pi_v = Real{3.141592653589793238462643383279502884};
template<typename T> constexpr T TEN(const int n) { return n == 0 ? T{1} : TEN<T>(n - 1) * T{10}; }
template<typename F> struct fix : F
{
    fix(F&& f) : F{std::forward<F>(f)} {}
    template<typename... Args> auto operator()(Args&&... args) const { return F::operator()(*this, std::forward<Args>(args)...); }
};
template<typename T, int n, int i = 0>
auto nd_array(int const (&szs)[n], const T x = T{})
{
    if constexpr (i == n) {
        return x;
    } else {
        return std::vector(szs[i], nd_array<T, n, i + 1>(szs, x));
    }
}
class printer
{
public:
    printer(std::ostream& os_ = std::cout) : m_os{os_} { m_os << std::fixed << std::setprecision(15); }
    template<typename... Args> int ln(const Args&... args) { return dump(args...), m_os << '\n', 0; }
    template<typename... Args> int el(const Args&... args) { return dump(args...), m_os << std::endl, 0; }
private:
    template<typename T> void dump(const T& v) { m_os << v; }
    template<typename T> void dump(const std::vector<T>& vs)
    {
        for (int i = 0; i < (int)vs.size(); i++) { m_os << (i ? " " : ""), dump(vs[i]); }
    }
    template<typename T> void dump(const std::vector<std::vector<T>>& vss)
    {
        for (int i = 0; i < (int)vss.size(); i++) { m_os << (0 <= i or i + 1 < (int)vss.size() ? "\n" : ""), dump(vss[i]); }
    }
    template<typename T, typename... Args> int dump(const T& v, const Args&... args) { return dump(v), m_os << ' ', dump(args...), 0; }
    std::ostream& m_os;
};
printer out;
class range
{
private:
    struct itr
    {
        itr(const int start = 0, const int step = 1) : m_cnt{start}, m_step{step} {}
        bool operator!=(const itr& it) const { return m_cnt != it.m_cnt; }
        int& operator*() { return m_cnt; }
        itr& operator++() { return m_cnt += m_step, *this; }
        int m_cnt, m_step;
    };
    int m_start, m_end, m_step;
public:
    range(const int start, const int end, const int step = 1) : m_start{start}, m_end{end}, m_step{step}
    {
        assert(m_step != 0);
        if (m_step > 0) { m_end = m_start + std::max(m_step - 1, m_end - m_start + m_step - 1) / m_step * m_step; }
        if (m_step < 0) { m_end = m_start - std::max(-m_step - 1, m_start - m_end - m_step - 1) / (-m_step) * (-m_step); }
    }
    itr begin() const { return itr{m_start, m_step}; }
    itr end() const { return itr{m_end, m_step}; }
};
range rep(const int end, const int step = 1) { return range(0, end, step); }
range per(const int rend, const int step = -1) { return range(rend - 1, -1, step); }
class scanner
{
public:
    scanner(std::istream& is_ = std::cin) : m_is{is_} { m_is.tie(nullptr), std::ios::sync_with_stdio(false); }
    template<typename T> T val()
    {
        T v;
        return m_is >> v, v;
    }
    template<typename T> T val(const T offset) { return val<T>() - offset; }
    template<typename T> std::vector<T> vec(const int n)
    {
        return make_v<T>(n, [this]() { return val<T>(); });
    }
    template<typename T> std::vector<T> vec(const int n, const T offset)
    {
        return make_v<T>(n, [this, offset]() { return val<T>(offset); });
    }
    template<typename T> std::vector<std::vector<T>> vvec(const int n0, const int n1)
    {
        return make_v<std::vector<T>>(n0, [this, n1]() { return vec<T>(n1); });
    }
    template<typename T> std::vector<std::vector<T>> vvec(const int n0, const int n1, const T offset)
    {
        return make_v<std::vector<T>>(n0, [this, n1, offset]() { return vec<T>(n1, offset); });
    }
    template<typename... Args> auto tup() { return std::tuple<std::decay_t<Args>...>{val<Args>()...}; }
    template<typename... Args> auto tup(const Args&... offsets) { return std::tuple<std::decay_t<Args>...>{val<Args>(offsets)...}; }
private:
    template<typename T, typename F>
    std::vector<T> make_v(const int n, F f)
    {
        std::vector<T> ans;
        for (int i = 0; i < n; i++) { ans.push_back(f()); }
        return ans;
    }
    std::istream& m_is;
};
scanner in;
int main()
{
    auto [N, K] = in.tup<int, int>();
    auto As = in.vec<int>(N);
    auto Bs = in.vec<int>(N);
    auto Pss = in.vvec<int>(N, N);
    using pii = std::pair<int, int>;
    auto calc = [&]() -> pii {
        auto Rs = nd_array<int>({N}, 0);
        for (const int i : rep(N)) {
            for (const int j : rep(N)) { Rs[i] += Pss[i][j]; }
        }
        auto Cs = nd_array<int>({N}, 0);
        for (const int j : rep(N)) {
            for (const int i : rep(N)) { Cs[j] += Pss[i][j]; }
        }
        int S = std::accumulate(Rs.begin(), Rs.end(), 0);
        auto xss = nd_array<ll>({N, N}, 0);
        int mi = -1, mj = -1;
        ll max = -inf_v<ll>;
        for (const int i : rep(N)) {
            for (const int j : rep(N)) {
                xss[i][j] = N * N * Pss[i][j] - N * (Rs[i] - As[i]) - N * (Cs[j] - Bs[j]) + (S - K);
                if (chmax(max, xss[i][j])) { mi = i, mj = j; }
            }
        }
        return {mi, mj};
    };
    for ([[maybe_unused]] const int _ : rep(K)) {
        const auto [i, j] = calc();
        As[i]--, Bs[j]--, Pss[i][j]--;
    }
    ll ans = 0;
    for (const int i : rep(N)) {
        for (const int j : rep(N)) { ans += (ll)Pss[i][j] * Pss[i][j]; }
    }
    out.ln(ans);
    return 0;
}