// ∞ Reasoning & Test-Time Compute
Test-Time Compute Scaling
Spending more inference compute — thinking, sampling, search — for better answers.
测试时计算扩展 2024 Advanced core
IN PLAIN WORDS
With the right strategy, spending compute at inference (verifier-guided search, revisions) can beat a 14× larger model. o1 showed performance rising smoothly with thinking time.
Used in: OpenAI o-series, DeepSeek-R1, Claude extended thinking
TIMELINE · prerequisites → this → what builds on it
01Original paper
Scaling LLM Test-Time Compute Optimally can be More Effective than Scaling Model Parameters
Charlie Snell, Jaehoon Lee, Kelvin Xu, et al. · 2024
arXiv ↗ ✓ Verified · arXiv API
02Where it sits
Learn the left column first; the right column builds on this one.
Learn first
Then you can learn
Related:DeepSeek-R1 Reasoning RL Recipe
Left: prerequisites · right: builds on this · dashed: related. Full map →
03Learning resources
Official docs, explainers, courses and reference code. All links checked.
04Where the course teaches it
On learning paths:Post-training & Reasoning Models
For AI / data
- Verification: verified, checked 2026-10-21.
- JSON: /api/learn/algorithms.json (id:
test-time-scaling)