jgrusewski 52a2663a2e feat(dqn): direction-branch Thompson sampling in experience_action_select ...

Replaces eps-greedy + Boltzmann on direction with:
  - Training: argmax of (sample_C51 + sample_IQN) per direction (Thompson)
  - Eval: argmax of (E_C51 + E_IQN) per direction (no exploration)

Conviction calculation now reads e_dir[] (joint E[Q] computed inline)
preserving the spec's requirement that conviction uses raw E[Q] not
samples (avoids Kelly cap jitter).

Magnitude/order/urgency branches: unchanged.

2026-04-29 16:22:43 +02:00

..

infra

feat(infra): GPU training launcher with local/cloud routing

2026-02-23 11:19:58 +01:00

plans

feat(dqn): enable Branching DQN with 45 factored actions (5×3×3)

2026-03-11 22:00:13 +01:00

superpowers

spec(dqn): Plan C Phase 2 ACTIVE — UCB asymmetry regression triggers Thompson resumption

2026-04-29 16:02:53 +02:00

dqn-gpu-hot-path-audit.md

perf(dqn): online softmax in compute_expected_q (kernel #3)

2026-04-28 23:33:14 +02:00

dqn-named-dims.md

fix(dqn): mag_concat_qdir SH2+b0 vs SH2+3 off-by-one — buffer + weight migration

2026-04-27 22:19:41 +02:00

dqn-wire-up-audit.md

feat(dqn): direction-branch Thompson sampling in experience_action_select

2026-04-29 16:22:43 +02:00

h_s2_consumers_audit.md

audit(dqn-v2): h_s2 audit anchor corrections (post-2c.3+4 reconnaissance)

2026-04-25 12:29:47 +02:00

health_diag_inventory.md

feat(diag): Phase 2A — health_diag_isv_mirror kernel + GpuHealthDiag orchestrator

2026-04-28 23:54:45 +02:00

isv-slots.md

refactor(dqn-v2): delete legacy CPU-DtoH per_branch_vsn_mean / per_branch_target_drift

2026-04-25 11:02:23 +02:00

lookahead-bias-audit-2026-04-28.md

fix(ofi): align OFI window to bar t's formation interval — kill 31/32-slot lookahead

2026-04-28 09:23:32 +02:00

ml-supervised-to-dqn-concept-audit.md

plan4(task7): Part E audit close-out — every supervised concept landed or OUT

2026-04-26 10:36:03 +02:00