Add v9 + summaries column; recommend running v9 bare (no --llm-summaries)

README.md

@@ -24,18 +24,20 @@ A fine-tuned code search embedding model based on **intfloat/e5-base-v2** (110M
 
 The headline results below are from cqs's production fixture — 218 queries (109 test + 109 dev) curated from real agent telemetry and LLM-generated retrieval cases on the cqs codebase itself. This is the eval that drives default-model decisions.
 
-| split | metric | BGE-large (1024-dim) | **v9-200k (768-dim)** | Δ |
-|-------|--------|---------------------:|----------------------:|--:|
-| test  | R@1    | 43.1%                | **45.9%**             | +2.8 |
-| test  | R@5    | 69.7%                | **70.6%**             | +0.9 |
-| test  | R@20   | **83.5%**            | 80.7%                 | -2.8 |
-| dev   | R@1    | 45.9%                | 46.8%                 | +0.9 |
-| dev   | R@5    | **77.1%**            | 68.8%                 | -8.3 |
-| dev   | R@20   | **86.2%**            | 81.7%                 | -4.5 |
+| split | metric | BGE-large (1024-dim) | **v9-200k bare (768-dim)** | v9-200k + LLM summaries |
+|-------|--------|---------------------:|---------------------------:|------------------------:|
+| test  | R@1    | 43.1%                | **45.9%**                  | 39.4%                   |
+| test  | R@5    | 69.7%                | **70.6%**                  | 69.7%                   |
+| test  | R@20   | **83.5%**            | 80.7%                      | 80.7%                   |
+| dev   | R@1    | 45.9%                | **46.8%**                  | 45.0%                   |
+| dev   | R@5    | **77.1%**            | 68.8%                      | 67.9%                   |
+| dev   | R@20   | 86.2%                | 81.7%                      | **86.2%**               |
 
 **v9-200k essentially ties BGE-large on test R@5 and edges it on R@1 across both splits** — at 1/3 the parameter count and 25% smaller embeddings (768 vs 1024 dim). The cost is on dev R@5/R@20 (~5–8 pp behind), where BGE-large's broader pre-training base helps on out-of-distribution queries. For latency-sensitive or memory-constrained workloads, v9-200k is the right choice.
 
-**Decision (2026-05-01):** cqs keeps BGE-large as default for the dev R@5 hedge, but v9-200k is shipped as a first-class opt-in preset. Set `CQS_EMBEDDING_MODEL=v9-200k` or `cqs slot create v9 --model v9-200k` to use it.
+**Run v9-200k bare — skip cqs's `--llm-summaries` enrichment pass for this model.** Adding LLM-generated summaries to chunks (and re-running the embedding pass over the summary-augmented text) *hurts* test R@1 by ~6 pp and is a wash on R@5 across both splits. The call-graph-trained dense channel already captures the signal summaries would add; injecting summary text dilutes the model's strongest top-1 signal. The only metric that materially benefits is dev R@20 (+4.5 pp). For BGE-large the same enrichment pass is a small net win; the v9-200k training distribution is the difference.
+
+**Decision (2026-05-01):** cqs keeps BGE-large as default for the dev R@5 hedge, but v9-200k is shipped as a first-class opt-in preset. Set `CQS_EMBEDDING_MODEL=v9-200k` or `cqs slot create v9 --model v9-200k` to use it. Don't pass `--llm-summaries` on the index command unless you're specifically optimizing for dev R@20.
 
 ## Historical results (296q synthetic fixture)