SYSTEM Cited by 1 source
SSR — Search Semantic Retriever (Meta)¶
SSR (Search Semantic Retriever) is the 12-layer, 200-million-parameter encoder model Meta uses to convert natural-language user queries into dense vector representations for the semantic-retrieval arm of Facebook Groups Scoped Search. Disclosed in the 2026-04-21 Meta Engineering post.
Role in the retrieval pipeline¶
- Pre-processed query arrives at SSR (post-tokenization/normalization/rewriting).
- SSR encodes the query into a dense vector in the same embedding space as the precomputed group-post index.
- Faiss performs approximate nearest neighbor search over the precomputed vector index of group posts.
- Semantic candidates flow into the L2 MTML ranker alongside Unicorn's lexical candidates, with cosine similarity scores as features.
The key contribution of SSR is resolving the natural-language-intent gap in keyword search:
"We needed a system where searching for an 'Italian coffee drink' effectively matches a post about 'cappuccino,' even if the word 'coffee' is never explicitly stated."
Model details¶
- Layers: 12
- Parameters: ~200 million
- Input: pre-processed user query (tokenized, normalized, rewritten)
- Output: dense vector in the group-post embedding space
- Training data, loss function, backbone family — not disclosed in the post
Relation to the broader Meta embedding-model portfolio¶
Not the same as EnCodec (audio), MTIA-served recsys embeddings (systems/meta-adaptive-ranking-model), or Llama — SSR is a purpose-built retrieval-encoder for scoped community search. Serves a narrower task: query → single dense vector for ANN lookup, not generation or ranking.