#pragma GCC optimize("Ofast,unroll-loops")
//#pragma GCC target("avx2,fma,bmi,bmi2,sse4.2")
#include<bits/stdc++.h>
#define pll pair<long long, long long>
#define fi first
#define se second
#define bit(i,j) ((j >> i) & 1)
#define lowbit(x) (x & (-x))
#define sigma main
using namespace std;
const long long inf = 1e18;
const int mod = 1e9+7;
const int MAXN = 1e6+100;
#define int long long
struct fenwick_tree{
#define ll long long
ll n;
vector<ll>ft;
void init(ll _n){
n = _n;
ft.resize(n + 10);
}
void add(ll idx, ll val){
while(idx <= n) ft[idx] = (ft[idx] + val), idx += (idx & (-idx));
}
ll get(ll idx){
ll res = 0;
while(idx > 0) res = (res + ft[idx]), idx -= (idx & (-idx));
return res;
}
ll get(ll l, ll r){return get(r) - get(l - 1);}
} T , S;
pll order[MAXN];
vector<int> adj[MAXN];
int Time = 0 , par[MAXN];
void dfs(int u , int p){
for(int i = 0 ; i < adj[u].size() ; i++){
int v = adj[u][i];
if(v == p) continue;
par[v] = u;
order[v].fi = ++Time;
dfs(v , u);
order[v].se = ++Time;
}
}
int32_t sigma(){
//freopen("MILITARYTRAINING.inp", "r", stdin);
//freopen("MILITARYTRAINING.out", "w", stdout);
ios_base::sync_with_stdio(0); cin.tie(0);
int n , q; cin >> n >> q;
for(int i = 1 ; i <= n ; i++){
int k; cin >> k;
while(k--){
int x; cin >> x;
adj[i].push_back(x);
}
}
for(int i = 1 ; i <= n ; i++){
reverse(adj[i].begin() , adj[i].end());
int t = adj[i].back();
adj[i].pop_back();
reverse(adj[i].begin() , adj[i].end());
adj[i].push_back(t);
}
dfs(1 , 1);
vector<pll> query;
vector<int> ans(q + 1);
for(int i = 1 ; i <= q; i ++){
int x; cin >> x;
query.push_back({x , i});
}
sort(query.begin() , query.end());
deque<pll> Q;
assert(Time == (n-1)*2);
T.init(Time+10) , S.init(Time+10);
Q.push_back({1 , 0});
int jj = 0;
int state = 0 , tl = 0 , cnt = 0;
while(Q.size()){
pll t = Q.front(); Q.pop_front();
int u = t.fi , st = t.se;
if(st > state){
for(; jj < query.size() && query[jj].fi <= tl + cnt * 2 ; jj++){
pll X = query[jj];
int res = -1;
int l = 1 , r = Time;
while(l <= r){
int mid = l + r >> 1;
if(S.get(1 , mid) >= X.fi - tl){
res = mid ; r = mid-1;
}else{
l = mid+1;
}
}
ans[X.se] = T.get(res , res);
}
tl += cnt * 2;
state++;
}
if(u != 1){
cnt++;
T.add(order[u].fi , u);
S.add(order[u].fi , 1);
T.add(order[u].se , par[u]);
S.add(order[u].se , 1);
}
vector<int> nxt;
int K = 0;
for(auto v : adj[u]){
if(v == par[u]) K++;
else{
if(K == 0 || nxt.size() == 0) nxt.push_back(v);
else Q.push_back({v , st + 1});
}
}
while(nxt.size() > 0){
Q.push_front({nxt.back() , st}); nxt.pop_back();
}
}
for(; jj < query.size() ; jj++){
pll &X = query[jj];
X.fi -= tl;
X.fi %= cnt * 2;
if(X.fi == 0) X.fi = cnt * 2;
int res = -1;
int l = 1 , r = Time;
while(l <= r){
int mid = l + r >> 1;
if(S.get(1 , mid) >= X.fi){
res = mid ; r = mid-1;
}else{
l = mid+1;
}
}
ans[X.se] = T.get(res , res);
}
for(int i = 1 ; i <= q ; i++) cout << ans[i] << "\n";
return 0;
}
