{ "benchmark": "AxiomicLabs/ArithMark-2.0", "checkpoint_step": 29000, "packed_tokens_at_checkpoint": 950272000, "parameters": 2901984, "scoring": "raw summed continuation log-likelihood", "precision": "float32", "overall": { "accuracy": 90.28, "correct": 2257, "total": 2500 }, "operator_count": { "1": 91.12, "2": 93.07, "3": 84.00 }, "single_operator_topics": { "addition": 97.21, "subtraction": 93.84, "multiplication": 72.92, "division": 76.92 }, "examples_containing_operator": { "+": 92.01, "-": 91.06, "*": 82.59, "/": 89.94 }, "topics": { "addition": 97.21, "division": 76.92, "mixed_three_ops": 88.84, "mixed_two_ops": 95.19, "multiplication": 72.92, "parentheses_three_ops": 79.46, "parentheses_two_ops": 90.70, "subtraction": 93.84 } }