jgrusewski
b0ac0afe63
test(ml): smoke tests for hierarchical softmax and diversity short-circuit
5 new tests:
- batch_hierarchical_softmax_actions_smoke: valid outputs, diversity, exposure coverage
- batch_hierarchical_softmax_low_temp_valid: edge case at near-zero temperature
- batch_hierarchical_vs_flat_softmax_diversity: hierarchical >= flat exposure levels
- eval_softmax_temp_floor_is_half: bounds [0.5, 2.0] enforced with clamp
- diversity_short_circuit_blocks_phantom_sharpe: 2/45 actions -> positive penalty
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
2026-03-05 17:30:24 +01:00
..
2026-03-01 22:47:39 +01:00
2026-03-05 00:22:04 +01:00
2026-03-05 00:22:04 +01:00
2026-03-01 22:47:39 +01:00
2026-03-01 22:47:39 +01:00
2026-03-01 22:47:39 +01:00
2026-03-01 22:47:39 +01:00
2026-03-05 17:30:24 +01:00
2026-03-01 22:47:39 +01:00
2026-03-01 22:47:39 +01:00
2026-03-01 22:47:39 +01:00
2026-03-01 22:47:39 +01:00
2026-03-01 22:47:39 +01:00
2026-03-05 00:18:58 +01:00
2026-03-01 22:47:39 +01:00
2026-03-03 22:16:35 +01:00