CCA-F
Start Free

Adaptive vs Fixed Decomposition

Advanced

Design task decomposition strategies for complex workflows · Difficulty 3/5

0%
decompositionadaptiveattention-dilutionarchitecture
Prerequisites
Task Decomposition & Routing Strategies

Explanation

Choosing between fixed and adaptive decomposition is a critical architectural decision that affects both quality and efficiency.

When to Use Fixed Pipelines

  • The workflow has well-defined stages (e.g., extract -> validate -> transform -> load)
  • Each stage has predictable inputs and outputs
  • The number of steps is known in advance
  • Quality comes from thoroughness at each stage, not exploration

    • When to Use Adaptive Decomposition

  • The scope is unknown until investigation begins
  • Early findings change what needs to be investigated next
  • Dependencies between subtasks are discovered during execution
  • The task requires exploration before planning

    • Attention Dilution

    Large inputs (e.g., reviewing a 50-file PR in one pass) cause attention dilution -- the model may miss issues in the middle of long contexts. The solution is decomposition:

  • Per-file local analysis (each file gets focused attention)
  • Cross-file integration pass (finds interactions between files)

    • This two-pass approach catches both local issues and cross-cutting concerns.

    Key Takeaways

    • Fixed pipelines for predictable workflows; adaptive decomposition for exploration
    • Attention dilution in large inputs requires decomposition into focused passes
    • Two-pass approach: per-file local analysis + cross-file integration

    Related Concepts

    Narrow Decomposition Risk

    Narrow task decomposition causes incomplete results even when all subtasks succeed

    Multi-Step Workflow Orchestration

    Decompose multi-concern requests and investigate in parallel

    Test Yourself

    1 / 2

    Your coordinator agent decomposes the research topic "Impact of AI on creative industries" into three subtasks: AI in digital art creation, AI in graphic design, and AI in photography. The final report appears comprehensive for visual arts but completely misses music, writing, and film. What's the root cause?