jgrusewski cd149cdb5d fix(rl): gate controller target 0.02→0.10, dead zone 0.5×/2× → 0.2×/5× ...

Controller oscillated: 3 dones at step 15k spiked EMA above 2×target,
triggering thousands of steps of tightening that killed trades.
Higher target (10% vs 2%) matches the natural trade frequency at
b=16. Wider dead zone (5× above / 0.2× below) prevents single-batch
spikes from triggering tighten/relax oscillation.

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

2026-05-25 10:43:09 +02:00

..

cuda

fix(rl): gate controller target 0.02→0.10, dead zone 0.5×/2× → 0.2×/5×

2026-05-25 10:43:09 +02:00

examples

feat(rl): exhaustive diag JSONL for all trade-management mechanics

2026-05-24 22:05:24 +02:00

src

fix(rl): gate controller target 0.02→0.10, dead zone 0.5×/2× → 0.2×/5×

2026-05-25 10:43:09 +02:00

tests

feat(rl): gate warmup — confidence + FRD gates inactive for first 10k steps

2026-05-25 10:16:05 +02:00

build.rs

feat(rl): adaptive gate threshold controller from trade frequency

2026-05-25 10:26:10 +02:00

Cargo.toml

feat(rl): R6 — replace LobEnv with kernel-driven GPU-pure env step

2026-05-23 10:39:25 +02:00