liveBook · Manning

Extractive summarization

As explained in Retrieval Augmented Generation, The Foundational Ideas

Extractive pruning reduces context by deleting selected material without generating replacement text.

Extractive summarization reduces context by deleting selected material without generating replacement text. Because it preserves original wording, it is a safer default when answers must cite source passages; abstractive compression can invent facts. section 13.9 section 13.4.1

Choose extractive pruning when the task rewards faithful retrieval of original text, then control its failure modes:

  • Use it for factual recall or answers that should reproduce a span from the source; prefer abstractive synthesis for summaries or multi-hop answers combining documents. section 13.7
  • Start around 4× compression, validate against uncompressed answers on held-out queries, and increase the ratio only while accuracy holds. section 13.7
  • Force-keep domain-critical terms such as drug names, contractual parties, model numbers, statute identifiers, and negations. section 13.7

Moderate compression versus over-pruning

Removing distractors can improve accuracy, but beyond roughly 6× compression, meaning increasingly disappears with the deleted tokens. A well-tuned system therefore trades under-extraction against over-pruning rather than maximizing compression. section 13.9 section 13.7

Table 13.4 Method tradeoffs
Compression paradigm
Primary methodology
Optimal ratio
Compute overhead
Risk of semantic loss
Heuristic Pruning
TF-IDF Filtering
Low (~2x)
Negligible
Moderate
Proxy Perplexity
LongLLMLingua
Moderate (4x - 10x)
Moderate
Low to Moderate
Native Pruning
AttentionRAG
Moderate (4x - 6x)
Varies
Very Low
Trained Synthesis
RECOMP, CORE
Extreme (10x - 30x)
High
Low
KV-Cache Eviction
H2O, SnapKV, RazorAttention
N/A (decoding-time)
Low
Very Low

Compare heuristic filtering, proxy-perplexity methods, native attention pruning, trained synthesis, and decoding-time cache eviction by ratio, overhead, and semantic-loss risk. section 13.7

How Retrieval Augmented Generation, The Foundational Ideas covers it

How deep it goes
Explains deletion-based selection, ratio tuning, force-keep terms, and failure modes, while contrasting pruning with synthesis and cache eviction.
Examples
Uses LongLLMLingua, AttentionRAG, RECOMP extractive, TF-IDF filtering, quoted clauses, and force-kept domain terminology as examples.
What you can do after
Apply it via compression-ratio evaluation, force-token constraints, and task-specific selection between extraction and synthesis.
How it connects
Connects extractive summarization to abstractive compression, retrieval-augmented generation, attention-based pruning, and KV-cache eviction.

Also explained in

Compare how 5 books explain extractive summarization →

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Retrieval Augmented Generation, The Foundational Ideas. The section links point to the parts of the book it draws on.