Files
gbrain/test/search/knobs-hash-reranker.test.ts
T
85286a556c fix(search): stop boosting compiled_truth at default detail (#3430) (#3514)
* fix(search): stop boosting compiled_truth at default detail (#3430)

COMPILED_TRUTH_BOOST = 2.0 is applied AFTER RRF normalization, and RRF's whole
dynamic range over a 100-deep pool is 1/60 -> 1/160 (a factor of 2.67). So a
2.0x multiplier consumes roughly three quarters of the range: break-even is
`2/(60+r) >= 1/60`, i.e. r <= 60, which means ANY boosted chunk inside the
first 60 ranks outranks an unboosted rank-1 chunk. That is a categorical
filter, not a tilt.

Measured against master's own rrfFusion, with the correct answer in a
fenced_code chunk at vector rank 0:

  compiled_truth chunks in pool | final rank | in top-20
  10                            | 10         | yes
  20                            | 20         | NO
  40                            | 40         | NO
  80                            | 59         | NO

With the boost off the answer stays at rank 0 in every case.

The gate was spelled `detail !== 'high'` -- written as though `high` were the
special case. The documented contract in src/core/operations.ts is
"low (compiled truth only), medium (default, all with dedup), high (all
chunks)", which makes LOW the special one: `low` already restricts to
compiled_truth, so a boost there is a no-op among equals, while `medium` and
`high` are both meant to see everything. So the default detail was silently
compiled-truth-only, contradicting the op's own description.

Three changes:

1. The three fusion call sites now route through a named predicate,
   `shouldBoostCompiledTruth(detail)`, returning true only for 'low'.
   Extracted rather than left inline precisely because an inline expression is
   only reachable through a full hybridSearch round trip -- which is why the
   inversion went unnoticed. The predicate is directly unit-testable.

2. KNOBS_HASH_VERSION 13 -> 14. Results are cached AFTER fusion, so rows
   ranked under the old semantics would otherwise be served under the new ones
   for the whole TTL (3600s default). One-time miss spike on upgrade.

3. test/search-compiled-truth-boost-scope.test.ts pins both the mapping and
   the arithmetic, and documents the displacement it prevents.

Verified the tests discriminate: stubbing the OLD predicate body into master
(so the failure is behavioral rather than a missing export) gives 4 fail /
3 pass; with the fix, 7 pass. typecheck clean, verify 32/32, and 144 pass /
0 fail across the search + fusion + cache suites.

* fix(test): update the three remaining KNOBS_HASH_VERSION pins to 14 (#3430)

Missed in the first pass because I ran a targeted set of test files instead of
the full suite. CI shards 3, 8 and 10 caught them:

  test/search/knobs-hash-reranker.test.ts:67
  test/cross-modal-phase1.test.ts:139,149
  test/search-alias-resolved-boost.test.ts:93

Each carries the running history of why the version moved, so each gets the
13→14 rationale appended rather than just the number swapped. No pins at 13
remain anywhere in test/.

---------

Co-authored-by: Garry Tan <garrytan@gmail.com>
2026-07-28 19:27:46 -07:00

250 lines
11 KiB
TypeScript

/**
* v0.35.0.0 — knobsHash reranker-field participation tests.
*
* Pins:
* - KNOBS_HASH_VERSION === 3 (bumped 1→2 v0.35.0.0 for reranker; 2→3 v0.35.6.0
* for floor_ratio — codex outside-voice T1 cross-floor cache contamination).
* - All 5 new reranker fields participate in the hash:
* reranker_enabled, reranker_model, reranker_top_n_in,
* reranker_top_n_out, reranker_timeout_ms.
* Each one flipping changes the hash → no two reranker configs share
* a cache row.
* - top_n_out=null vs unset shows up as 'none' in the hash (no NaN).
* - Append-only convention (CDX2-F13): the existing 9 fields hash
* identically under v=2 as they did under v=1 for the same input
* when the reranker section is held constant. Reordering them
* would silently rebuild the hash for every existing row.
* - Mid-deploy invariant (CDX2-F12): the v=1 prefix in the hash input
* differs from the v=2 prefix; a tokenmax v=1 process and a v=2
* process produce distinct row IDs for the same (source_id, query).
*/
import { describe, test, expect } from 'bun:test';
import {
knobsHash,
KNOBS_HASH_VERSION,
resolveSearchMode,
MODE_BUNDLES,
type ResolvedSearchKnobs,
} from '../../src/core/search/mode.ts';
import { resolveHardExcludes } from '../../src/core/search/source-boost.ts';
/** Build a baseline resolved knob set with all reranker fields filled. */
function baseKnobs(): ResolvedSearchKnobs {
return {
...MODE_BUNDLES.balanced,
reranker_enabled: false,
reranker_model: 'zeroentropyai:zerank-2',
reranker_top_n_in: 30,
reranker_top_n_out: null,
reranker_timeout_ms: 5000,
resolved_mode: 'balanced',
mode_valid: true,
};
}
describe('KNOBS_HASH_VERSION + version invariants', () => {
test('version is 13 (…; 10→11 asymmetric input_type #1400; 11→12 hard-excludes #2825; 12→13 embedding-provider migration #3390)', () => {
// v0.35.0.0: 1→2 to fold reranker fields. v0.35.6.0: 2→3 to fold
// floor_ratio. v0.36 wave: piggybacks on v=3 with 7 cross-modal knobs
// (D2) PLUS column + provider context (D8/CDX-2 cross-column isolation).
// v0.40.4 (salem) + v0.39 T21 (master): 3→4 to fold graph_signals AND
// schema_pack name + version (graph-on cache write cannot be served to
// graph-off; cross-pack contamination structurally impossible).
// v0.40.3.0 (D8): 4→5 to fold contextual_retrieval + kill switch,
// sequenced behind salem's v=4 graph-signals.
// v0.41.22.0 (type-unification): 5→6 to fold the alias_resolved
// post-fusion boost. Cache rows written before the boost stage
// cannot leak past the new stage. T2: 6→7 title_boost. v0.42.3.0: 7→8
// autocut. issue #1777: 8→9 archive/ demote (search-exclude policy change
// isn't in the hash, so the bump invalidates archive-excluded cache rows).
// v0.43: 9→10 relational recall arm (rel=/reld=).
// #1400: 10→11 asymmetric input_type fix — embedQuery() now produces
// query-side vectors for asymmetric providers, so rows keyed on
// pre-fix document-side query vectors must not be served.
// #2825: 11→12 to fold the resolved hard-exclude prefix list (hx=) —
// cached rows leaked GBRAIN_SEARCH_EXCLUDE'd slugs across processes.
// #3430: 13→14 — the compiled_truth boost no longer applies at
// detail=medium. Results are cached after fusion, so rows ranked under
// the old boost semantics must not be served under the new ones.
expect(KNOBS_HASH_VERSION).toBe(14);
});
test('hash is 16 hex chars regardless of reranker config', () => {
const a = knobsHash(baseKnobs());
const b = knobsHash({ ...baseKnobs(), reranker_enabled: true });
expect(a).toMatch(/^[0-9a-f]{16}$/);
expect(b).toMatch(/^[0-9a-f]{16}$/);
});
});
describe('Each reranker field flips the hash (cache-row separation)', () => {
test('reranker_enabled false vs true → different hash', () => {
const off = knobsHash({ ...baseKnobs(), reranker_enabled: false });
const on = knobsHash({ ...baseKnobs(), reranker_enabled: true });
expect(off).not.toBe(on);
});
test('reranker_model differs → different hash', () => {
const z2 = knobsHash({ ...baseKnobs(), reranker_model: 'zeroentropyai:zerank-2' });
const z1 = knobsHash({ ...baseKnobs(), reranker_model: 'zeroentropyai:zerank-1' });
const z1s = knobsHash({ ...baseKnobs(), reranker_model: 'zeroentropyai:zerank-1-small' });
expect(new Set([z2, z1, z1s]).size).toBe(3);
});
test('reranker_top_n_in differs → different hash', () => {
const a = knobsHash({ ...baseKnobs(), reranker_top_n_in: 30 });
const b = knobsHash({ ...baseKnobs(), reranker_top_n_in: 50 });
expect(a).not.toBe(b);
});
test('reranker_top_n_out null vs 10 → different hash', () => {
const noTrunc = knobsHash({ ...baseKnobs(), reranker_top_n_out: null });
const trunc10 = knobsHash({ ...baseKnobs(), reranker_top_n_out: 10 });
expect(noTrunc).not.toBe(trunc10);
});
test('reranker_timeout_ms differs → different hash (CDX2-F14)', () => {
// CDX2-F14: a timeout change (5s → 100ms) changes search behavior
// (more fail-opens) so stale cache rows must invalidate. Without
// this field in parts[], the rows would silently match.
const t5 = knobsHash({ ...baseKnobs(), reranker_timeout_ms: 5000 });
const t1 = knobsHash({ ...baseKnobs(), reranker_timeout_ms: 1000 });
expect(t5).not.toBe(t1);
});
});
describe('mid-deploy invariant (CDX2-F12)', () => {
test('tokenmax-with-reranker vs tokenmax-without-reranker → distinct hashes', () => {
// tokenmax mode bundle has reranker on. An operator who flips it off
// via `gbrain config set search.reranker.enabled false` produces a
// different cache row, not a shared one.
const tokenmaxOn = knobsHash(resolveSearchMode({ mode: 'tokenmax' }));
const tokenmaxOff = knobsHash(resolveSearchMode({
mode: 'tokenmax',
overrides: { reranker_enabled: false },
}));
expect(tokenmaxOn).not.toBe(tokenmaxOff);
});
test('conservative vs balanced vs tokenmax → 3 distinct hashes', () => {
const c = knobsHash(resolveSearchMode({ mode: 'conservative' }));
const b = knobsHash(resolveSearchMode({ mode: 'balanced' }));
const t = knobsHash(resolveSearchMode({ mode: 'tokenmax' }));
expect(new Set([c, b, t]).size).toBe(3);
});
});
describe('determinism + stability', () => {
test('same input → same hash (re-call)', () => {
const k = baseKnobs();
expect(knobsHash(k)).toBe(knobsHash(k));
});
test('same mode bundle → same hash across resolveSearchMode calls', () => {
const a = knobsHash(resolveSearchMode({ mode: 'balanced' }));
const b = knobsHash(resolveSearchMode({ mode: 'balanced' }));
expect(a).toBe(b);
});
test('top_n_out=null renders as "none" in parts[] (no NaN)', () => {
// CDX2-F15 + F16 + F14 collide here. The parts[] line is
// `rro=${knobs.reranker_top_n_out ?? 'none'}` — null must produce a
// stable string token, never `NaN` or `null`.
const h1 = knobsHash({ ...baseKnobs(), reranker_top_n_out: null });
const h2 = knobsHash({ ...baseKnobs(), reranker_top_n_out: null });
expect(h1).toBe(h2);
expect(h1).toMatch(/^[0-9a-f]{16}$/);
});
});
describe('append-only convention (CDX2-F13)', () => {
test('parts[] order in source: reranker fields appear AFTER the existing 9', async () => {
const src = await Bun.file(
new URL('../../src/core/search/mode.ts', import.meta.url),
).text();
// Locate the parts[] declaration. The existing 9 fields end with
// `lim=${knobs.searchLimit}`. The 5 new fields must appear AFTER
// that line. Reordering would silently rebuild the hash for every
// existing v=2 cache row.
const limIdx = src.indexOf('lim=${knobs.searchLimit}');
const rrIdx = src.indexOf('rr=${knobs.reranker_enabled');
expect(limIdx).toBeGreaterThan(0);
expect(rrIdx).toBeGreaterThan(0);
expect(rrIdx).toBeGreaterThan(limIdx);
});
test('v=3 additions: col= and prov= appear AFTER the reranker block', async () => {
// v0.36 D8: cache-key contamination across embedding columns + providers.
// The two new tokens must sit at the bottom of parts[] so existing v=2
// hashes can only differ in those positions — keeping the append-only
// chain auditable for future v=4 readers.
const src = await Bun.file(
new URL('../../src/core/search/mode.ts', import.meta.url),
).text();
const rrtIdx = src.indexOf('rrt=${knobs.reranker_timeout_ms');
const colIdx = src.indexOf('col=${ctx?.embeddingColumn');
const provIdx = src.indexOf('prov=${ctx?.embeddingModel');
expect(rrtIdx).toBeGreaterThan(0);
expect(colIdx).toBeGreaterThan(rrtIdx);
expect(provIdx).toBeGreaterThan(colIdx);
});
test('v=3 fields participate: column flip changes the hash', () => {
const k = baseKnobs();
const defaultCol = knobsHash(k, { embeddingColumn: 'embedding', embeddingModel: 'openai:text-embedding-3-large' });
const voyageCol = knobsHash(k, { embeddingColumn: 'embedding_voyage', embeddingModel: 'voyage:voyage-3-large' });
expect(defaultCol).not.toBe(voyageCol);
});
test('v=3 fields participate: same column + different provider → different hash', () => {
const k = baseKnobs();
const a = knobsHash(k, { embeddingColumn: 'embedding', embeddingModel: 'openai:text-embedding-3-large' });
const b = knobsHash(k, { embeddingColumn: 'embedding', embeddingModel: 'openai:text-embedding-3-small' });
expect(a).not.toBe(b);
});
test('v=3 fields fall back to embedding/default when ctx undefined', () => {
// Backward-compat: callers that don't know the column (e.g. telemetry
// helpers) should still produce a stable hash matching the default
// 'embedding' + 'default' provider pair.
const k = baseKnobs();
const bare = knobsHash(k);
const explicit = knobsHash(k, { embeddingColumn: 'embedding', embeddingModel: 'default' });
expect(bare).toBe(explicit);
});
});
describe('v=12 hard-exclude participation (#2825)', () => {
test('different exclude lists → different hashes', () => {
const k = baseKnobs();
const noEnv = knobsHash(k, { hardExcludes: resolveHardExcludes(undefined, undefined, undefined) });
const withEnv = knobsHash(k, { hardExcludes: resolveHardExcludes(undefined, undefined, 'private/') });
expect(noEnv).not.toBe(withEnv);
});
test('include (opt-back-in) changes the hash too', () => {
const k = baseKnobs();
const a = knobsHash(k, { hardExcludes: resolveHardExcludes(undefined, undefined, undefined) });
const b = knobsHash(k, { hardExcludes: resolveHardExcludes(undefined, ['test/'], undefined) });
expect(a).not.toBe(b);
});
test('same prefixes in different input order → SAME hash (normalization)', () => {
const k = baseKnobs();
const a = knobsHash(k, { hardExcludes: ['a/', 'b/', 'test/'] });
const b = knobsHash(k, { hardExcludes: ['test/', 'b/', 'a/'] });
expect(a).toBe(b);
});
test('undefined hardExcludes is stable (legacy-caller fallback)', () => {
const k = baseKnobs();
expect(knobsHash(k)).toBe(knobsHash(k));
// ...and distinct from an explicit resolved default list — a legacy
// caller can never collide with a policy-carrying cache row.
expect(knobsHash(k)).not.toBe(
knobsHash(k, { hardExcludes: resolveHardExcludes(undefined, undefined, undefined) }),
);
});
});