| {"schema": 2, "epoch": 87617, "nonce": "5dac468bda6bc562", "hotkey": "5Hg4fML2RQ86kiurzDJsHXBMwJuCDzW6Jv4MU8zTeLjonNxH", "source_hash": "24837b9ae6895829747c5eb448673693fedbeb2e3e62a5ca170051a512053fc0", "weights_hash": "f105aad63d7b96494c90b5ac1f2a4e3fa0b94f9a6a7430370730cc1f32438ded", "model_id": "router", "total_cost_usd": 0.006424930000000001, "n_calls": 6, "call_log_hash": "a243e6bb1fc4861dd0cedc7e2394a9c4baaf9e955a726a715a87aa293d014a06", "measurement": "1449fadb4821cadef93f7eecc8c3b040e2cd244e01a2607ea531e5f7055c38d8", "confined": true, "latency_s": 313.152, "tokens_in": 2297, "tokens_out": 46917, "results": [{"benchmark": "mmlu", "task_id": "mmlu-12053", "answer": "B", "cost_usd": 2.9e-05, "chosen_rung": 4, "rungs_used": [4], "distribution": [0.150576, 0.117191, 0.07348, 0.07348, 0.413866, 0.072902, 0.098506]}, {"benchmark": "mmlu", "task_id": "mmlu-5882", "answer": "C", "cost_usd": 3.490000000000001e-05, "chosen_rung": 4, "rungs_used": [4], "distribution": [0.11633, 0.129196, 0.075098, 0.075098, 0.397995, 0.091932, 0.11435]}, {"benchmark": "math", "task_id": "gsm8k-903", "answer": "1. Small masks: \\(20 \\div 4 = 5\\) groups, requiring \\(5 \\times 2 = 10\\) yards.\n2. Large masks: \\(8 \\div 3 = \\frac{8}{3}\\) groups, requiring \\(\\frac{8}{3} \\times 2.25 = 6\\) yards.\n3. Total: \\(10 + 6 = 16\\)\n\n\\[\n\\boxed{16}\n\\]", "cost_usd": 0.00010890000000000002, "chosen_rung": 4, "rungs_used": [4], "distribution": [0.06871, 0.086161, 0.066134, 0.066134, 0.562166, 0.067442, 0.083252]}, {"benchmark": "math", "task_id": "gsm8k-784", "answer": "- Daily round-trip distance: \\(5 \\times 2 = 10\\) miles \n- Weekly distance: \\(10 \\times 5 = 50\\) miles \n- Fuel used: \\(50 \\div 10 = 5\\) gallons \n- Weekly cost: \\(5 \\times \\$3 = \\$15\\)\n\n15", "cost_usd": 9.459999999999997e-05, "chosen_rung": 4, "rungs_used": [4], "distribution": [0.095289, 0.093462, 0.071449, 0.071449, 0.481236, 0.091754, 0.09536]}, {"benchmark": "code", "task_id": "lcb-abc400_b", "answer": "n, m = map(int, input().split())\n\nlimit = 10**9\ntotal = 0\npower = 1\n\nfor _ in range(m + 1):\n total += power\n if total > limit:\n print(\"inf\")\n break\n power *= n\nelse:\n print(total)\n", "cost_usd": 7.14e-05, "chosen_rung": 4, "rungs_used": [4], "distribution": [0.104071, 0.108677, 0.074204, 0.074204, 0.345218, 0.145565, 0.148061]}, {"benchmark": "code", "task_id": "lcb-abc400_d", "answer": "```python\nimport sys\nfrom collections import deque\n\n# Increase recursion depth just in case, though not strictly needed here\nsys.setrecursionlimit(2000)\n\ndef solve():\n # Fast I/O\n input_data = sys.stdin.read().split()\n if not input_data:\n return\n\n iterator = iter(input_data)\n \n try:\n H = int(next(iterator))\n W = int(next(iterator))\n \n S = []\n for _ in range(H):\n S.append(next(iterator))\n \n A = int(next(iterator)) - 1\n B = int(next(iterator)) - 1\n C = int(next(iterator)) - 1\n D = int(next(iterator)) - 1\n except StopIteration:\n return\n\n # Distance matrix initialized with infinity\n INF = 10**9\n dist = [[INF] * W for _ in range(H)]\n \n dq = deque()\n dq.append((A, B))\n dist[A][B] = 0\n \n # Localize deque methods for slight performance gain\n dq_append = dq.append\n dq_appendleft = dq.appendleft\n dq_popleft = dq.popleft\n \n while dq:\n r, c = dq_popleft()\n d = dist[r][c]\n \n # If we reached the target, since it's 0-1 BFS, this is the shortest path.\n if r == C and c == D:\n print(d)\n return\n\n # --- 0-weight moves (Adjacent Roads) ---\n # Move to adjacent cell if it is a road ('.')\n # Cost 0. Push to front.\n \n # Up\n nr, nc = r - 1, c\n if nr >= 0:\n if S[nr][nc] == '.':\n if dist[nr][nc] > d:\n dist[nr][nc] = d\n dq_appendleft((nr, nc))\n \n # Down\n nr, nc = r + 1, c\n if nr < H:\n if S[nr][nc] == '.':\n if dist[nr][nc] > d:\n dist[nr][nc] = d\n dq_appendleft((nr, nc))\n \n # Left\n nr, nc = r, c - 1\n if nc >= 0:\n if S[nr][nc] == '.':\n if dist[nr][nc] > d:\n dist[nr][nc] = d\n dq_appendleft((nr, nc))\n \n # Right\n nr, nc = r, c + 1\n if nc < W:\n if S[nr][nc] == '.':\n if dist[nr][nc] > d:\n dist[nr][nc] = d\n dq_appendleft((nr, nc))\n \n # --- 1-weight moves (Kicks) ---\n # Kick in 4 directions. Affects dist 1 and dist 2.\n # Cost 1. Push to back.\n next_d = d + 1\n \n # Up\n # Dist 1\n nr, nc = r - 1, c\n if nr >= 0:\n if dist[nr][nc] > next_d:\n dist[nr][nc] = next_d\n dq_append((nr, nc))\n \n # Dist 2\n nr, nc = r - 2, c\n if nr >= 0:\n if dist[nr][nc] > next_d:\n dist[nr][nc] = next_d\n dq_append((nr, nc))\n \n # Down\n # Dist 1\n nr, nc = r + 1, c\n if nr < H:\n if dist[nr][nc] > next_d:\n dist[nr][nc] = next_d\n dq_append((nr, nc))\n \n # Dist 2\n nr, nc = r + 2, c\n if nr < H:\n if dist[nr][nc] > next_d:\n dist[nr][nc] = next_d\n dq_append((nr, nc))\n \n # Left\n # Dist 1\n nr, nc = r, c - 1\n if nc >= 0:\n if dist[nr][nc] > next_d:\n dist[nr][nc] = next_d\n dq_append((nr, nc))\n \n # Dist 2\n nr, nc = r, c - 2\n if nc >= 0:\n if dist[nr][nc] > next_d:\n dist[nr][nc] = next_d\n dq_append((nr, nc))\n \n # Right\n # Dist 1\n nr, nc = r, c + 1\n if nc < W:\n if dist[nr][nc] > next_d:\n dist[nr][nc] = next_d\n dq_append((nr, nc))\n \n # Dist 2\n nr, nc = r, c + 2\n if nc < W:\n if dist[nr][nc] > next_d:\n dist[nr][nc] = next_d\n dq_append((nr, nc))\n\nif __name__ == '__main__':\n solve()\n```", "cost_usd": 0.00608613, "chosen_rung": 0, "rungs_used": [0], "distribution": [0.310156, 0.132798, 0.0659, 0.0659, 0.150019, 0.153059, 0.122169]}], "quote": {"measurement": "1449fadb4821cadef93f7eecc8c3b040e2cd244e01a2607ea531e5f7055c38d8", "report_data": "d4e82540b97e7c3feb30c41c20cecc2dd35b92012db6e4817a28a1abaed341dc", "platform_sig": "tdx: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"}} |