01 Reweight 3,408
Paths that led to the real fix get stronger. Paths that misled it fade.
Ran in this pass00 / A graph with a memory
Fixes learned1281django/django
Held-out issue → past fix → gold function
Measured path, drawn schematically.
For platform teams running coding agents on mature codebases.
Start an unfamiliar bug investigation with ranked functions and the graph evidence behind them. Feed back the confirmed fix so the next investigation can use it.
Strict Function Acc@10
319 retrospective Django fix rows
All gold functions must appear in the top ten. This is one recorded Django snapshot, which includes later fixes. Three test rows share tickets with training; excluding them preserves the gain. Read the evaluation audit.
Recorded Django evidence. The route is schematic; the function, rank and supporting fix come from the measured replay.
Scope of evidence
01 / The learning pass
Predict. See the real fix. Learn.
Then ask the same questions again.
1281 fixes learned in time order. The same 319 held-out fixes scored at every checkpoint. Three test rows share tickets with training; this is a retrospective replay on a fixed code snapshot.
File Acc@1 rises from 41.07% to 44.51%. Function Acc@10 rises from 45.14% to 48.28%. The learned graph beats dense on both. MRR ends at 0.430. Lines connect measured checkpoints, including the dips; baselines are horizontal references.
Paths that led to the real fix get stronger. Paths that misled it fade.
Ran in this passA direct road for a repeated detour.
Ran in this passA missing link, added when it cannot reach the fix.
Ran in this passResolve shared names and divided meanings.
Ran in this pass| Fixes learned | RepoCortex | Frozen | Dense | BM25 |
|---|---|---|---|---|
| 0 | 41.07% | 41.07% | 35.74% | 34.48% |
| 92 | 41.69% | 41.07% | 35.74% | 34.48% |
| 183 | 41.69% | 41.07% | 35.74% | 34.48% |
| 274 | 41.69% | 41.07% | 35.74% | 34.48% |
| 366 | 41.38% | 41.07% | 35.74% | 34.48% |
| 458 | 41.69% | 41.07% | 35.74% | 34.48% |
| 549 | 42.01% | 41.07% | 35.74% | 34.48% |
| 640 | 42.32% | 41.07% | 35.74% | 34.48% |
| 732 | 42.32% | 41.07% | 35.74% | 34.48% |
| 824 | 42.32% | 41.07% | 35.74% | 34.48% |
| 915 | 42.63% | 41.07% | 35.74% | 34.48% |
| 1006 | 42.63% | 41.07% | 35.74% | 34.48% |
| 1098 | 44.2% | 41.07% | 35.74% | 34.48% |
| 1190 | 45.14% | 41.07% | 35.74% | 34.48% |
| 1281 | 44.51% | 41.07% | 35.74% | 34.48% |
| Fixes learned | RepoCortex | Frozen | Dense | BM25 |
|---|---|---|---|---|
| 0 | 45.14% | 45.14% | 37.3% | 32.92% |
| 92 | 44.83% | 45.14% | 37.3% | 32.92% |
| 183 | 44.51% | 45.14% | 37.3% | 32.92% |
| 274 | 44.83% | 45.14% | 37.3% | 32.92% |
| 366 | 44.83% | 45.14% | 37.3% | 32.92% |
| 458 | 45.14% | 45.14% | 37.3% | 32.92% |
| 549 | 45.77% | 45.14% | 37.3% | 32.92% |
| 640 | 46.08% | 45.14% | 37.3% | 32.92% |
| 732 | 46.08% | 45.14% | 37.3% | 32.92% |
| 824 | 46.08% | 45.14% | 37.3% | 32.92% |
| 915 | 46.39% | 45.14% | 37.3% | 32.92% |
| 1006 | 46.39% | 45.14% | 37.3% | 32.92% |
| 1098 | 46.71% | 45.14% | 37.3% | 32.92% |
| 1190 | 48.28% | 45.14% | 37.3% | 32.92% |
| 1281 | 48.28% | 45.14% | 37.3% | 32.92% |
The frozen reference is this pass's step 0 graph, with learning off. Learning gain is measured from that starting point.
On 319 chronological fix rows, function Acc@10 is 41.69% for dense + BM25 without graph, 45.14% for the frozen graph, and 48.28% after learning. All gold functions must appear in the top ten.
This is a retrospective replay on one Django snapshot. Three test rows share tickets with training. Excluding them leaves 316 rows: 42.09%, 45.57%, and 48.73%, respectively. The snapshot includes later fixes; these are offline results, not live request measurements.
Read the evaluation audit · Download the reproduced comparison
A measured gain on 319 held-out fix rows. Not a claim about every codebase.
02 / The evidence
Each candidate has a route through the code graph. Inspect a recorded path, the agent study, or the hosted search check.
A past issue and its learned FIXED_IN link lead to the candidate. This is a captured replay result.
held-out issue #36605│└─ issue #35690 / similarity 0.74 └─ FIXED_IN / weight 0.90 └─ django.db.models.query.QuerySet.in_bulk└─ concept:bulk -SHORTCUT-> django.db.models.query.QuerySet.in_bulk (hits 3, misses 1)└─ concept:list -SHORTCUT-> django.db.models.query.QuerySet.in_bulk (hits 3, misses 1) rank 1 / score 0.825file / django/db/models/query.pygold match / yes gold functions└─ func:django/db/models/query.py:django.db.models.query.QuerySet.in_bulk└─ func:django/db/models/query.py:django.db.models.query.QuerySet.in_bulk.get_obj
Similarity, edge weight, and ranking score are distinct values. Across 319 held-out fix rows, 314 hits used FIXED_IN paths and 63 used SHORTCUT paths; these are hit counts, not accuracy.
Inspect search results →63,885 → 53,888 mean input tokens per issue. Asking the agent to call the tool first reduced input tokens by 27%.
Measured with claude-sonnet-5-5. One run per condition; input totals include cached reads. This study measures investigation, not successfully completed fixes.
Read the agent study →30 Django tasks. Top-1 correctness: 13/30 with the tool, 12/30 without. No established correctness gain.
Four concurrent requests against the hosted app, with live HydraDB traversal. Median response 3.41 s; p95 3.69 s, including network and queueing.
Inspect the request results →Four repeated reports. A small read-only smoke check, not sustained capacity, cold-start, or learning-write load.