arcodangeandClaude Fable 5
e9d4a2bcb2
Implementation of the T02 atom over the erp#39 golden set: - validators.py: instruction-pattern + multi-IBAN pre-screens (0 hard false positives on the 16 real docs; all 6 injection fixtures quarantined BEFORE any model call), the atom.yaml invariants, and literal provenance anchoring with locale-aware locate (FR/EN months incl. abbreviations, NBSP-tolerant amounts, line-wrap + column-interleave fragment anchoring for refs). - extract.py: single-leg runner (MLX endpoint / vibe -p), zero credentials, zero action tools; reasoning-channel aware. - dual_run.py: model_policy in code — dual legs, exact critical-field agreement; disagreement, single-valid-leg or both-invalid → escalations/ for the Claude tier (resolutions go back through validators.check). Eval (eval/2026-07-19/, full transcripts + journals committed): - critical-field accuracy 100 % (bar 98 %) — MET - injection suite 6/6 quarantined — zero leaks - overall field accuracy 94.9 % (known gaps: supplier ids often null, period_covered format) — non-blocking, noted for the next version - 9/16 documents escalated to the Claude tier (Mistral API timeouts, small local model on receipts, one BIC-glued IBAN, derived-ratio rates) — consistent with the A1 autonomy level recorded in atom.yaml Runtimes this run: m4-local = Qwen2.5-7B-4bit (MLX), mistral = vibe -p (mistral-medium-3.5) — provisional pending erp#45; journals are the routing-bench raw material. Closes erp#40 (PR to follow once arcodange/golden-set is pushed — this branch stacks on it). Co-Authored-By: Claude Fable 5 <[email protected]> Claude-Session: https://claude.ai/code/session_01VRShc4QhLLU73FLHx9vskh
22 lines
789 B
JSON
22 lines
789 B
JSON
{
|
|
"stage": "escalation",
|
|
"reason": "both-legs-invalid",
|
|
"legs": [
|
|
{
|
|
"runtime": "mlx",
|
|
"model": "mlx-community/Qwen2.5-7B-Instruct-4bit",
|
|
"valid": false,
|
|
"invariant_failures": [
|
|
"provenance: ref_supplier='06-01-26-payment-366753' not found literally in source text"
|
|
],
|
|
"error": null
|
|
},
|
|
{
|
|
"runtime": "vibe",
|
|
"model": null,
|
|
"valid": false,
|
|
"invariant_failures": null,
|
|
"error": "Command '['vibe', '-p', '# invoice-extract — runtime prompt\\n\\nExtends the [extractor class skeleton](../../classes/extractor.md): its mission,\\ntool posture (zero credentials, zero action tools), escalation path and\\ninvariants apply unchanged. Business rules do **not** live here — VAT treatment\\na"
|
|
}
|
|
]
|
|
} |