AppliedAIPrep logoAppliedAI/Prep
← LLM & GenAI Fundamentals / 38
hardOpenAIDeepSeekGoogle

How are reasoning models (o1/R1-style) trained, and what is test-time compute scaling?

Reasoning models are the 2025-2026 frontier. The signal is knowing they are trained (largely via RL on verifiable rewards) to produce long internal reasoning, and that they scale on a second axis: spending more compute at inference.

Updated Aug 2026 · Grounded in real Applied AI Engineer interview loops and written to a senior-engineer editorial bar.

Reasoning models are the 2025-2026 frontier. The signal is knowing they are trained (largely via RL on verifiable rewards) to produce long internal reasoning, and that they scale on a second axis: spending more compute at inference.

Unlock the other 750 answers · ₹2,000 / $25Your progress and mastery stay saved · 6 months · one payment · no auto-renew
UP NEXT ON YOUR JOURNEY
Next in this trackHow does LLM tool calling (function calling) actually work under the hood?Next in this trackHow do RoPE and ALiBi encode position, and why do they extrapolate better than learned positions?Next in this trackWhat are Multi-Query (MQA) and Grouped-Query Attention (GQA), and why do they exist?
DISCUSSION · 0

No comments yet — be the first to share your approach.