Codex db8e6d11f2 malkhut(scoring): fast scalar + advantage mode, reward execution quality ...

Fast scalar mode (default, for CMA loop):
  - Rewards: fill quality (PnL when fills happen), moderate fill rate (5-15% sweet spot)
  - Tolerates: no-fills (valid advisory recommendation)
  - Penalizes: extreme fill rates (<3% lazy, >30% picked off), adverse selection, drawdown
  - Light noop penalty (-0.5) vs old heavy (-50) — no-fills are valid signals

Advantage mode (for offline analysis):
  - advantage = raw_performance - baseline_performance
  - baseline = exponential moving average (decay=0.995)
  - Clipped to [-10, +10]
  - Reduces score variance 5.5x vs raw scoring

Scoring mode selection:
  PolicyEvaluator(scoring_mode='fast') — default for CMA loop
  PolicyEvaluator(scoring_mode='advantage') — for offline analysis

8 new tests for scoring modes. Total: 1186 tests, 50 files, all green.

2026-07-13 13:38:32 +02:00

..

__init__.py

malkhut(T6): training core — CMA-ES trainer, registry, pipeline, selector

2026-07-11 10:33:56 +02:00

_DATA_STORAGE_SCHEMA_FORMATS.md

malkhut: three-layer identifier architecture for cross-system asset identification

2026-07-12 00:40:14 +02:00

advantage_scorer.py

malkhut(scoring): fast scalar + advantage mode, reward execution quality

2026-07-13 13:38:32 +02:00

asset_behavior.py

malkhut: asset classification, behavior DSL, auto-compiler, behavior-driven scenarios

2026-07-11 06:37:17 +02:00

asset_bridge.py

malkhut: asset bridge — directory ↔ classification integration

2026-07-12 10:45:37 +02:00

asset_classification.py

malkhut: three-layer identifier architecture for cross-system asset identification

2026-07-12 00:40:14 +02:00

asset_compiler.py

malkhut: asset classification, behavior DSL, auto-compiler, behavior-driven scenarios

2026-07-11 06:37:17 +02:00

cma_trainer.py

malkhut(scoring): fast scalar + advantage mode, reward execution quality

2026-07-13 13:38:32 +02:00

cognition.py

malkhut(T8): cognition pipeline + regime expansion + prod tooling

2026-07-11 10:39:03 +02:00

discrepancy.py

malkhut(T4): Strategy DSL v2 + generator + supporting modules

2026-07-11 10:28:38 +02:00

dsl.py

malkhut(T4): Strategy DSL v2 + generator + supporting modules

2026-07-11 10:28:38 +02:00

execution_quality.py

malkhut(T4): Strategy DSL v2 + generator + supporting modules

2026-07-11 10:28:38 +02:00

generator.py

malkhut(T4): Strategy DSL v2 + generator + supporting modules

2026-07-11 10:28:38 +02:00

hooks.py

malkhut(T4): Strategy DSL v2 + generator + supporting modules

2026-07-11 10:28:38 +02:00

importance.py

malkhut(T4): Strategy DSL v2 + generator + supporting modules

2026-07-11 10:28:38 +02:00

monitor.py

malkhut(T8): cognition pipeline + regime expansion + prod tooling

2026-07-11 10:39:03 +02:00

news_sources.py

malkhut(T8): cognition pipeline + regime expansion + prod tooling

2026-07-11 10:39:03 +02:00

observability.py

malkhut(T4): Strategy DSL v2 + generator + supporting modules

2026-07-11 10:28:38 +02:00

parallel_eval.py

malkhut(perf): parallel episode eval — 9x single-eval speedup, zero fidelity loss

2026-07-11 20:19:32 +02:00

parallel.py

malkhut(T4): Strategy DSL v2 + generator + supporting modules

2026-07-11 10:28:38 +02:00

pipeline.py

malkhut(T6): training core — CMA-ES trainer, registry, pipeline, selector

2026-07-11 10:33:56 +02:00

ray_eval.py

malkhut(optim): vectorized reward + Ray parallel eval + VBT post-analysis

2026-07-12 23:56:16 +02:00

regime_expansion.py

malkhut(T8): cognition pipeline + regime expansion + prod tooling

2026-07-11 10:39:03 +02:00

registry.py

malkhut(T6): training core — CMA-ES trainer, registry, pipeline, selector

2026-07-11 10:33:56 +02:00

rollback.py

malkhut(T4): Strategy DSL v2 + generator + supporting modules

2026-07-11 10:28:38 +02:00

selector.py

malkhut(T6): training core — CMA-ES trainer, registry, pipeline, selector

2026-07-11 10:33:56 +02:00

stress.py

malkhut(T4): Strategy DSL v2 + generator + supporting modules

2026-07-11 10:28:38 +02:00

structured_obs.py

malkhut(T4): Strategy DSL v2 + generator + supporting modules

2026-07-11 10:28:38 +02:00

trajectory.py

malkhut(T4): Strategy DSL v2 + generator + supporting modules

2026-07-11 10:28:38 +02:00

vbt_analysis.py

malkhut(optim): vectorized reward + Ray parallel eval + VBT post-analysis

2026-07-12 23:56:16 +02:00