Reasoning & Test-Time Compute · 2022
Solving Quantitative Reasoning Problems with Language Models (Minerva)
Showed that a large model trained on mathematical and scientific text, with chain-of-thought and majority voting, could solve quantitative reasoning problems far better than prior models without external tools.
Editorial record
Plain-language summary
Minerva continues pretraining a large language model on math-heavy web and arXiv data, then uses step-by-step prompting and samples many solutions to vote on the answer. It substantially raised scores on MATH and STEM problem sets using only the model’s own reasoning. It marked how far scaled models plus test-time sampling could push mathematical reasoning.
Knowledge graph
Relationships
Descendants
ExtendsEvidence: Strongly supported
Chain-of-Thought Prompting Elicits Reasoning
Minerva extends chain-of-thought to quantitative and mathematical reasoning
P-522
Provides evaluation forEvidence: Direct
GSM8K / MATH (mathematical reasoning)
Minerva is evaluated on the MATH benchmark
P-522
Source record
Provenance
- Record ID
- P-522
- Record created
- 2026-07-13
- Last reviewed
- 2026-07-14
- Record version
- 2
- https://arxiv.org/abs/2206.14858
- arXiv:2206.14858
Citation caveat: Citation metadata is approximate and marked unverified in the source dataset.