# Stage 146: Score 256D Pattern Tensor (aq146_score_256d_pattern_tensor / DB-UNIT-04)

Embeds architectural patterns, components, and repositories into dense 256-dimension [0..10000] integer tensors across 8 semantic 32-axis blocks.

- Atomic micro-queue: `aq146_score_256d_pattern_tensor` (visibility: `public_commons`)
- Backing database unit: `DB-UNIT-04` (`48-Cell Hybrid Search, LSH & Retrieval Core`)
- Database cell: `lsh-db.teleon.dev` (`teleon_core.teleon_lsh_signatures_256`, `teleon_core.bm25f_seven_field_postings`, `teleon_core.roaring_trigram_inverted_index`, `teleon_core.markov_l0_prefetch_transitions`)
- Claim row hydration: `single_self_contained_flat_task_row` (100% context inline, normalized writeback)
- Track 2 Sister queue (`SISTER-04`): `sq04-improve-search-queue.teleon.dev` (Tune BM25F field weights, MMR diversity, synonym rings, multilingual stemmers, and 0..N retrieval steps.)
- Lanes: SMALLINT 256D 0..10000, 8x32 Semantic Blocks, 0-Call Structural Hash

## Execution engines & AI Agent freedom contract

- **Deterministic (0 model calls)**: `bm25f_roaring_lsh_zero_model_v1` (8-block 256D integer tensor encoder & cosine/L1 projector (0 calls))
- **Hybrid (<=1 model call)**: `query_intent_expansion_hybrid_v1`
- **AI Agent & Bot Fleet (1..6 calls)**: `multi_hop_federated_scout_v1` (Deep architectural trade-off & cross-domain axis calibrator)
  - Contract: **Instructions + Resources + Freedom to search for new resources and new ways not bound by instructions**

## Database invariants

- Retrieval stays deterministic at 0 model calls when optimize_mode=zero_model_deterministic.
- Return compact contract cards first; defer source bytes until inspect_or_stream.

## External & internal bot fleet actions

- Quantize 8 32-dimension blocks (determinism, resilience, finops, security, composability, latency, ux, welfare) into [0..10000]
- Support sub-millisecond integer dot-product and LSH neighbor lookups without float drift
- Track tensor trajectory shifts across component revisions

## Sister self-tuning areas

- Tune 7-field BM25F weights and MMR diversity_floor on frozen retrieval benchmarks
- Expand multilingual query synonym tables and domain aspect boosters
- Add pre-retrieval / post-retrieval deterministic reranking steps and code_blocks