BiteSizedChunks.comLearn one small thing at a time.

Course contentsShow

585What is an AI Agent?
586The Perception-Reasoning-Action Loop
587Observation Space and Input Processing
588Reasoning and Decision Making
589Action Space and Tool Calling
590Loop Termination Conditions
591Iteration Limits and Safeguards
592Synchronous vs Asynchronous Execution
593Error Handling in the Agent Loop
594Logging and Observability for Agent Loops
595What Is Agent Memory?
596Short-Term vs Long-Term Memory
597Memory Types: Semantic, Episodic, Procedural
598In-Context Memory via Prompts
599Memory Summarization Techniques
600Vector Memory for Semantic Retrieval
601Entity Memory and Knowledge Graphs
602Memory Indexing and Retrieval Strategies
603Memory Write Operations and Updates
604Forgetting and Memory Pruning
605Memory Consistency and Conflicts
606Multi-User Memory Isolation
607Planning vs Reactive Agent Behavior
608Single-Step vs Multi-Step Planning
609Task Decomposition Fundamentals
610Plan-and-Execute Architecture
611ReAct Planning Pattern
612Goal Stack Planning
613Hierarchical Task Networks
614Replanning and Plan Repair
615Beam Search and Plan Ranking
616Dynamic Replanning Triggers
617Plan Verification and Validation
618Planning Budget and Depth Limits
619Agent State: What to Track
620State Persistence Strategies
621State Serialization and Checkpointing
622Stopping Conditions: Max Iterations
623Stopping Conditions: Goal Achievement
624Stopping Conditions: Error and Timeout Handling
625State Pruning and Memory Management
626Resumable Agents and Long-Running Tasks

585What is an AI Agent?
586The Perception-Reasoning-Action Loop
587Observation Space and Input Processing
588Reasoning and Decision Making
589Action Space and Tool Calling
590Loop Termination Conditions
591Iteration Limits and Safeguards
592Synchronous vs Asynchronous Execution
593Error Handling in the Agent Loop
594Logging and Observability for Agent Loops
595What Is Agent Memory?
596Short-Term vs Long-Term Memory
597Memory Types: Semantic, Episodic, Procedural
598In-Context Memory via Prompts
599Memory Summarization Techniques
600Vector Memory for Semantic Retrieval
601Entity Memory and Knowledge Graphs
602Memory Indexing and Retrieval Strategies
603Memory Write Operations and Updates
604Forgetting and Memory Pruning
605Memory Consistency and Conflicts
606Multi-User Memory Isolation
607Planning vs Reactive Agent Behavior
608Single-Step vs Multi-Step Planning
609Task Decomposition Fundamentals
610Plan-and-Execute Architecture
611ReAct Planning Pattern
612Goal Stack Planning
613Hierarchical Task Networks
614Replanning and Plan Repair
615Beam Search and Plan Ranking
616Dynamic Replanning Triggers
617Plan Verification and Validation
618Planning Budget and Depth Limits
619Agent State: What to Track
620State Persistence Strategies
621State Serialization and Checkpointing
622Stopping Conditions: Max Iterations
623Stopping Conditions: Goal Achievement
624Stopping Conditions: Error and Timeout Handling
625State Pruning and Memory Management
626Resumable Agents and Long-Running Tasks

← AI Engineering

Lesson 615 of 1,886·15. AI Agents - ArchitecturePro lesson

Beam Search and Plan Ranking

Generating multiple candidate plans and selecting the most promising based on heuristics or cost estimates.

This lesson is for subscribers

You've completed the free preview. Subscribe to unlock every lesson in every course.

See pricing Back to course