Semantic memory

As explained in AI Agents in Action

Semantic memory is a memory form organized around meanings, including conceptual information and associated feelings or emotions.

Semantic memory stores meanings: global or local facts, concepts, preferences, and associated feelings or emotions. In an AI agent, it provides long-term information that can be retrieved when relevant. section 8.6 section 8.6.2

A semantic memory workflow typically:

• Captures facts, preferences, or other memory-bearing statements. section 8.6.2
• Processes input semantically and extracts relevant questions for retrieval. section 8.6.2
• Retrieves relevant memories from a vector database to augment an agent prompt. section 8.6 section 8.9

One statement, two memory interpretations

The same statements can populate different memory forms. A semantic store converts them into the relevant meaning-based memory type, while episodic and procedural stores emphasize events or steps. section 8.6.2

Figure 8.17 How semantic memory augmentation works

Trace how semantic augmentation transforms input into queries for more relevant memory retrieval. section 8.6.2

How AI Agents in Action covers it

How deep it goes

Explains semantic categorization, query extraction, retrieval, and configuration, but does not specify customizable memory prompts.

Examples

Uses facts and preferences, semantic memory stores, and comparisons of identical statements represented as different memory types.

What it assumes

Assumes basic awareness of agent memory and retrieval while distinguishing sensory, short-term, and long-term memory.

What you can do after

Operate/configure semantic memory stores and understand how semantic augmentation improves relevant agent recall.

How it connects

Connects semantic memory with long-term memory, episodic memory, procedural memory, retrieval-augmented generation, and vector databases.

liveBookmanning.com

Read AI Agents in Action on liveBook Get the book at manning.com

Also explained in

• 

  liveBookmanning.com
  AI Agents in Action, Second EditionSemantic memory is memory retrieved according to the meaning of stored information rather than only its explicit graph relationships or literal terms.

Compare how 2 books explain semantic memory →

Related concepts

• memorykind of · Memory is a storage resource or mechanism that retains data or computational state for access during computation or later use.
• hybrid memory systemused by · A hybrid memory system combines semantic retrieval with graph-based retrieval and updates to support an agent's memory.
• AI agentpart of · An AI agent is a computational system that uses artificial intelligence to interpret information, select actions, and pursue goals through interaction with an environment.
• vector databaseuses · A vector database is a data-management system that stores and indexes numerical vector representations and supports retrieval through similarity or distance searches.

This explainer was generated by Manning's knowledge graph from the text of AI Agents in Action. The section links point to the parts of the book it draws on.