AI / LlamaIndex Interview Questions
What are the response modes available in LlamaIndex query engines?
The response_mode parameter controls how a response synthesizer turns retrieved Nodes into a final answer, and LlamaIndex offers several depending on how much thoroughness versus speed you need.
| Mode | Behavior |
refine | Processes Nodes one at a time, updating the answer sequentially with one LLM call per Node |
compact | Packs as many Nodes as fit into the context window per call, reducing LLM calls versus refine |
tree_summarize | Recursively summarizes Nodes bottom-up like a tree until one final answer remains |
simple_summarize | Truncates and stuffs Nodes into a single prompt in one shot |
accumulate | Generates a separate answer per Node and concatenates them, without merging |
compact is the default for most query engines because it balances answer quality against LLM call count reasonably well.
Which response mode processes Nodes sequentially, one LLM call per Node?
The default response mode for most LlamaIndex query engines is:
More Related questions...
