Reasoning & Test-Time Compute · 2025
s1: Simple Test-Time Scaling
Shows strong o1-style reasoning can be reproduced from 1,000 curated traces plus 'budget forcing' (appending 'Wait') to control think-time, making test-time scaling cheap and transparent.
Editorial record
Plain-language summary
Shows strong o1-style reasoning can be reproduced from 1,000 curated traces plus 'budget forcing' (appending 'Wait') to control think-time, making test-time scaling cheap and transparent.
Knowledge graph
Relationships
Descendants
Conceptual ancestorEvidence: Strongly supported
Chain-of-Thought Prompting Elicits Reasoning
Builds on the reasoning lineage in the archive
freshness sweep 2026
Source record
Provenance
- Record ID
- P-636
- Record created
- 2026-07-13
- Last reviewed
- 2026-07-14
- Record version
- 2
- https://arxiv.org/abs/2501.19393
- arXiv:2501.19393
Citation caveat: Citation metadata is approximate and marked unverified in the source dataset.