liveBook · Manning

Reasoning pattern

As explained in Build a DeepSeek Model (From Scratch)

A reasoning pattern is a recurring strategy for carrying out multi-step inference or problem solving.

A reasoning pattern is a recurring strategy for multi-step inference or problem solving. In language-model training, models can acquire such patterns through reinforcement learning, synthetic examples, or distillation rather than relying only on explicit human instruction. section 1.2.3

Reasoning patterns emerge through a staged training process:

  • Begin with a capable pretrained base and a small cold-start dataset. section 1.2.3
  • Use pure reinforcement learning to explore problem-solving strategies through trial and error. section 1.2.3
  • Generate multiple candidate responses, retain stronger outputs through rejection sampling, and reuse them as synthetic training data. section 1.2.3
  • Blend synthetic outputs with supervised examples before applying final reinforcement learning across diverse tasks and formats. section 1.2.3

The base model determines how readily a pattern is absorbed. A model with prior mathematical knowledge can learn mathematical reasoning more effectively, while smaller models may repeat text instead of sustaining long reasoning chains. section 8.7.3

Distilling reasoning into smaller models

A large teacher model generates high-quality reasoning data for smaller student models. Distilled checkpoints from 1.5B to 70B parameters demonstrate the practical trade-off between inherited reasoning capability and computational efficiency. section 1.2.3

Figure 1.6 The multi-step post-training pipeline used to create DeepSeek-R1 from the DeepSeek-V3 base model. This process involves a combination of reinforcement learning (Pure RL), data generation (Rejection sampling), and fine-tuning to instill advanced reasoning capabilities.

Trace the multi-step pipeline from a lightly fine-tuned base through pure reinforcement learning, rejection sampling, data blending, and final reinforcement learning. section 1.2.3

How Build a DeepSeek Model (From Scratch) covers it

How deep it goes
Traces a five-stage training pipeline and explains base-model effects, but stops short of specifying reinforcement-learning algorithms or objective functions.
Examples
Uses DeepSeek-R1 training stages, distilled 1.5B–70B checkpoints, mathematical reasoning, and repetition in smaller models as examples.
What it assumes
Assumes familiarity with pretrained base models, reinforcement learning, supervised examples, and teacher-student knowledge distillation.
What you can do after
Evaluate how base-model knowledge, reinforcement learning, and distillation affect the acquisition and transfer of reasoning strategies.
How it connects
Links reasoning patterns to reinforcement learning, rejection sampling, supervised fine-tuning, knowledge distillation, and model compression.

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Build a DeepSeek Model (From Scratch). The section links point to the parts of the book it draws on.