> ## Documentation Index
> Fetch the complete documentation index at: https://wiki.platelunchcollective.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Post-Training

> Post-training refers to the processes applied to a foundation model after initial pre-training — including fine-tuning on task-specific data, reinforcement l...

*Technical implementation* · *AI Search Infrastructure*

## Definition

Post-training refers to the processes applied to a foundation model after initial pre-training — including fine-tuning on task-specific data, reinforcement learning from human feedback (RLHF), and instruction tuning. Post-training shapes how a model responds to queries, follows instructions, and prioritizes different types of information.

## Why It Matters for AI Search

Post-training is why two models trained on similar data can behave very differently when asked about a brand. Fine-tuning and RLHF introduce biases, priorities, and behavioral patterns that affect citation behavior, tone, and the weighting of different source types. A brand optimized for one model's base knowledge may not perform equally well across models with different post-training configurations.

## Related Terms

<CardGroup cols={2}>
  <Card title="Foundation model" href="/ai-search-glossary/foundation-model" />

  <Card title="Pre-training" href="/ai-search-glossary/pre-training" />

  <Card title="Fine-tuning" href="/ai-search-glossary/fine-tuning" />

  <Card title="Inference" href="/ai-search-glossary/inference" />

  <Card title="Training corpus" href="/ai-search-glossary/training-corpus" />
</CardGroup>

## Relevant Plate Lunch Collective Services

[AI SEO](https://www.platelunchcollective.com/services/ai-seo)  [AI Search Visibility Assessment](https://www.platelunchcollective.com/services/context-map)