Context-preserving multi-granularity embeddings
Parent: Concept knowledge architecture strategies · Published reference · snapshot 2026-10-22
↓ Facts as markdownall context files
Source-anchored research for Context-preserving multi-granularity embeddings. Run status: BUDGET_EXHAUSTED.
These notes link each claim to its source. Some citations identify archived research reports that are no longer hosted on this site. A published reference means the content is available; it does not certify independent review or accuracy.Read the editorial policy and follow the sources before relying on a claim.
How it works
Comparisons and alternatives
- Anthropic reports limited gains from generic document summaries added to chunks, unlike its chunk-specific context. (low confidence) [source]
Facts and statements
- Voyage contextualized embedding accepts grouped chunks so each chunk is encoded with its document peers. (low confidence) (verified as of 2026-10-22) [source]
- Voyage contextualized embedding preserves chunk order in the returned document embeddings. (low confidence) (verified as of 2026-10-22) [source]
- Voyage supports query and document input types that add task-specific prompts before embedding. (low confidence) (verified as of 2026-10-22) [source]
- Anthropic contextual retrieval prepends chunk-specific explanatory context before embedding and BM25 indexing. (low confidence) [source]
- Late chunking contextualizes document tokens before pooling token spans into individual chunk vectors. (low confidence) [source]
- Dense X defines propositions as minimal facts that include the context required to interpret them independently. (low confidence) [source]
- Late chunking can be applied to suitable long-context embedding models without additional training. (low confidence) [source]
- Dense X reports retrieval and QA improvements from proposition units under its evaluated computation budgets. (low confidence) [source]
- Chunk boundaries, size and overlap remain empirical design choices in Anthropic's contextual retrieval recipe. (low confidence) [source]
- Dense X's evaluated retrieval corpus is Wikipedia, limiting direct evidence about other corpora. (low confidence) [source]
- The late-chunking paper's overlap ablation does not show a clear retrieval advantage from overlapping chunks. (low confidence) [source]
- Dense X's manual proposition audit found both unsupported propositions and propositions lacking standalone context. (low confidence) [source]
Children
- Hierarchical summary retrieval (frontier)
- Late interaction token retrieval (frontier)
- Contextualized embedding model training (frontier)
Frontier under this node: Contextualized embedding model training, Hierarchical summary retrieval, Late interaction token retrieval