#include <bits/stdc++.h>
using i64 = long long;
constexpr i64 mod = 1000000007;
i64 add(i64 a, i64 b) {
return (a + b) % mod;
}
i64 sub(i64 a, i64 b) {
return (a + mod - b) % mod;
}
i64 mul(i64 a, i64 b) {
return (a * b) % mod;
}
void eadd(i64 &a, i64 b) {
a = add(a, b);
}
void esub(i64 &a, i64 b) {
a = sub(a, b);
}
void emul(i64 &a, i64 b) {
a = mul(a, b);
}
i64 calcPaint(i64 x, i64 y) {
return mul((1 + x) * x / 2 % mod, (1 + y) * y / 2 % mod);
}
int main() {
int N;
std::cin >> N;
std::vector<std::pair<i64, i64>> P(N);
for (int i = 0; i < N; ++i) {
std::cin >> P[i].first;
}
for (int i = 0; i < N; ++i) {
std::cin >> P[i].second;
}
std::vector<i64> rW(N + 1);
for (int i = 0; i < N; ++i) {
rW[i + 1] = rW[i] + P[i].second;
}
auto calcRangeWidth = [&](const int l, const int r) {
return rW[r] - rW[l];
};
std::map<i64, std::vector<i64>> eV;
for (int i = 0; i < N; ++i) {
eV[P[i].first].push_back(i);
}
std::set<std::tuple<i64, i64, i64>> rangeS; // (first, length)
rangeS.insert({0, calcRangeWidth(0, N), 0});
i64 answer = 0;
for (const auto &[h, vec] : eV) {
for (const int i : vec) {
const auto p = calcRangeWidth(0, i);
auto itr = rangeS.lower_bound({p + 1, 0, 0});
assert(itr != rangeS.begin());
--itr;
// eval
const auto [l, width, mh] = *itr;
i64 a = calcPaint(h % mod, width % mod), b = calcPaint(mh % mod, width % mod);
esub(a, b);
eadd(answer, a);
rangeS.erase(itr);
i64 lw = p - l, rw = (l + width) - (p + P[i].second);
if (lw != 0) {
rangeS.insert({l, lw, h});
}
if (rw != 0) {
rangeS.insert({p + P[i].second, rw, h});
}
}
}
std::cout << answer << std::endl;
}
