Post-mortem of the H6 Phase 1 smoke (WR = 50.21% verdict in
`docs/dqn-wire-up-audit.md`) traced an actual pearl violation in the
H6 implementation itself: state slot 121 uses the [0, 1] range
(`aux_softmax[env, 1] = p_up`) with sentinel 0.5, while every OTHER
state slot uses 0 as the "no signal" baseline (zero-padding,
feature_mask, ofi-missing, mtf-missing).
Per `pearl_first_observation_bootstrap`: "sentinel = 0; first
observation replaces directly." The H6 design violated this for
slot 121 alone. The encoder must learn TWO things about slot 121:
(1) the directional mapping AND (2) the appropriate bias offset for
the non-zero baseline — every other dim is single-step (mapping only).
Phase 2 is the simplest possible amplification fix: rewrite the bridge
to use the same convention as every other state slot. If the encoder
can't gradient-couple even after this fix, H6 is truly falsified and
we pivot to amplitude scaling or deeper hypothesis.
Change scope (atomic per `feedback_no_partial_refactor`):
- aux_softmax_to_per_env_kernel.cu: write `2*p_up - 1` instead of `p_up`
- gpu_experience_collector.rs: cold-start + FoldReset sentinel 0.5 → 0.0
- experience_kernels.cu: NULL-fallback in 3 state-gather kernels
0.5f → 0.0f
- state_layout.rs / state_layout.cuh: comment updates to reflect
[-1, +1] range and 0 sentinel
Estimated effort: ~45 min walltime (edits + verification gates) +
~25–40 min smoke wall-clock.
Verdict criteria (cycle 1 WR + a_var [d/m/o/u] in HEALTH_DIAG[0]):
- WR > 50.5% → recentering binding, H6 + Phase 2 sufficient → justify A2
- a_var for mag/ord/urg > 1e-3 → sub-branches learning under recentered signal
- WR pinned at 50.1–50.2% → Phase 2 falsified, pivot to amplitude
scaling or deeper hypothesis
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Two new plan docs to enable clean session handoff:
1. docs/plans/2026-05-12-sp22-h6-aux-policy-state-bridge.md
Detailed 10-step implementation runbook for H6 (aux→policy state
bridge). Each step has file paths, kernel signatures, expected
diff, and risk callouts. Estimated ~5hr for Phase 1 (training-side
+ A3 eval fallback). A2 (eval-side aux integration) is +1 day
follow-up if H6 confirmed.
2. docs/plans/2026-05-12-sp22-h6-next-session-prompt.md
Concise context-loading prompt to paste into next session.
Includes state of the world, runbook pointer, verification gates,
and start-here pointer.
Why staged: H6 implementation crosses kernel-level state-layout
contract (every consumer of STATE_DIM must migrate atomically per
feedback_no_partial_refactor). Doing it RIGHT requires careful
incremental verification with compute-sanitizer between steps —
not safely batched into a tail-end session.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
H1 result (commit 9adbca826, smoke train-5zmkr, reverted at e8814079d):
aux_dir_acc HD[2] = 78% with H=200 — aux head LEARNS direction at
longer horizon. But policy WR stayed pinned at 50.1-50.2% — the
learned aux signal does NOT propagate to policy (per
pearl_separate_aux_trunk_when_shared_starves: aux on separate
trunk with stop-grad to policy).
H6 design (synthesized from all v5-v11 + H1 evidence):
Wire the aux head's directional probability into the policy STATE
as an input feature.
- Policy gradient flows THROUGH the feature (uses it)
- Stop-grad blocks gradient BACK (aux trunk unaffected, pearl
preserved)
- Uses existing padding slot [121..128) in STATE_DIM=128 (no
layout growth)
Why this is the structural fix:
- Aux PROVED directional signal is in the features (78% at H=200)
- Policy PROVED it can't extract direction (WR=50% across all
v5-v11 conditions)
- Bridge connects the two without violating trunk separation
- Information-theoretic: gives policy a feature it provably
can't compute itself
Test outcome interpretation:
WR > 50.5% → Mechanism 1 was binding (trunk separation gap)
WR pinned → Mechanism 2 (reward density) or Mechanism 3 (V/A
unidentifiability) dominates → H3 or V/A fix next
Files changed:
- docs/plans/2026-05-12-sp22-wr-plateau-investigation.md:
H6 added as new primary hypothesis after H1; experiment order
revised
- docs/dqn-wire-up-audit.md: H6 design entry with three-mechanism
synthesis
Implementation scope (separate commit):
1. State layout: claim slot in padding [121..128) for aux_dir_prob
2. Aux trunk export: pull "up" probability per bar from aux forward
3. Experience collection: write aux_dir_prob into per-bar state
4. Stop-grad verification: confirm policy gradient blocked
5. Trade-open persistence: latch aux_dir_prob for trade duration
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
SP21 T2.2 status: cascade wiring COMPLETE.
- All E1-E8 enrichment producers wired to consumers
- Local compute-sanitizer clean on eval-baseline closure path
- v9 cycle 1 confirms cascade live: win_conc=1.72, curric_conc=0.31,
hindsight_mag=1.47e-5
- Eval pipeline now hard-fails on GPU error, no CPU fallback,
factored-action branch sizes + default ISV all wired
v10 hypothesis test (10 epochs, killed at E8 with clear answer):
- val_Sharpe peaks at E3 (174), monotonically degrades to E8 (66, -62%)
- WR pinned at 50.1-50.2% across ALL 8 epochs (no improvement)
- PF pinned at 1.00-1.01
- Cascade controllers stable but cannot move policy off the plateau
- Verdict: 3-epoch baseline structure IS optimal; longer training
monotonically degrades. WR plateau is upstream of the cascade.
Implication: SP21 T2.2 cascade is necessary but insufficient for
project_goal_wr_55_pf_2 (WR≥55%, PF≥2.0). Need new SP arc to address
the WR=50% plateau at its source.
New plan: docs/plans/2026-05-12-sp22-wr-plateau-investigation.md
Hypotheses (priority order):
H1: Label horizon mismatch (cheapest, most likely)
H2: Action-space pathology (Hold hiding directional signal)
H3: Reward shape (no directional gradient)
H4: Feature representation gap (MTF features zero-padded)
H5: Bar resolution itself is too noisy (longshot)
Each hypothesis tested as atomic smoke with one variable changed
vs v9 baseline (peak val=174, WR=50.1%).
Phase 1 milestone: WR ≥ 51% on val for ≥ 1 fold.
Phase 2: WR ≥ 53% across 3 folds.
Phase 3: WR ≥ 55% AND PF ≥ 2.0 (SP20+ goal achieved).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Adds the continuation plan for SP21 T2.2 Phase 1.5 (entry_q tracking
in portfolio_state slot 6) + Phase 2 (wire enrichment to real
per-trade tape from gpu_backtest_evaluator). Self-contained plan
that a fresh session can pick up cold:
- State-at-session-start summary with all 8 prior commits
- Phase 1.5 design (storage slot, capture site, plumbing,
EvalTrade extension)
- Phase 2 design (training_loop wire-up, enrichment refactor,
E5 alternate-signal options with recommendation)
- Hard rules carried from prior session (no NULLs, no stubs,
atomic per feedback_no_partial_refactor)
- Verification plan (5 GPU oracle test suites)
- Open design question (E5 alternate signal) with recommendation
- Multi-phase continuation map (Phases 3-7 + Phase 8)
- Final cascade verification (smoke run criteria)
Also amends the parent SP21 plan
(2026-05-10-sp21-train-eval-coherence-isv-defrost.md) with the
T2.2 multi-phase scope section that documents the full per-trade-tape
expansion the user committed to (option 3 — full wiring across
multiple sessions instead of the original recommended option (b)
aggregate-stats refactor).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Closes the Phase 3.2 forward-reference loop in the SP20 aggregator.
Previously `out_inputs->per_bar_hold_reward = 0.0f` was hardcoded; the
per-bar Hold opportunity-cost producer existed
(experience_kernels.cu:3823 — `per_bar_opp_cost = -aux_conf × cost_scale`)
and wrote to `hold_baseline_buffer` and `r_micro` directly, but never
reached HOLD_REWARD_EMA. Result: HOLD_REWARD_EMA frozen at sentinel 0.0
across all observed epochs; the SP20 reward centering loop
(`r_micro += per_bar_opp_cost - HOLD_REWARD_EMA`) stayed uncentered,
biasing the policy's reward signal away from zero-mean.
T3.3 wireup mirrors the T2.2 alpha refactor: a per-env scratch buffer
that the producer writes at every step (alongside the existing
`hold_baseline_buffer` write), and the aggregator reads at the same
step. Sums opp_cost over Hold-direction envs only; emits the mean as
`per_bar_hold_reward`. The HOLD_REWARD_EMA's gate (`hold_fraction > 0.5f`
from T3.2) preserves the strict-majority semantic from before.
Atomic across producer site, kernel signature, aggregator, launcher,
collector, and tests (per feedback_no_partial_refactor):
- experience_kernels.cu — new `float* per_bar_opp_cost_per_env`
kernel arg, NULL-tolerant write at line ~3823.
- sp20_aggregate_inputs_kernel.cu — new arg, 6th shmem stripe
(`sh_opp_cost_sum`), per-thread Hold-gated accumulation, tree
reduction extended to 6 stripes, output `per_bar_hold_reward =
opp_cost_sum / hold_count` when hold_count > 0 else 0.
- sp20_aggregate_inputs.rs — launcher signature, dynamic_shmem_bytes
5→6 stripes, doc table, internal test.
- gpu_experience_collector.rs — new `per_bar_opp_cost_per_env:
CudaSlice<f32>` field, alloc, struct construction, kernel-arg
pass at both experience_env_step and sp20_aggregate_inputs
launches (raw_ptr).
- sp20_aggregate_inputs_test.rs — helper renamed
`run_kernel_with_is_win_and_opp_cost`, 4 existing sites pass
NULL, 2 NEW oracle tests verifying real producer + NULL fallback.
- sp20_phase1_4_wireup_test.rs — NULL fallback at the wireup site;
HOLD_REWARD_EMA-stays-at-zero assertion remains valid.
Verification:
- cargo check -p ml --tests: passes (warnings only)
- cargo test -p ml --test sp20_aggregate_inputs_test --features cuda
-- --ignored: 12/12 GPU oracle tests pass on RTX 3050 Ti, including
both new T3.3 tests (per_bar_hold_reward_means_over_hold_envs_only,
null_per_bar_opp_cost_emits_zero).
- cargo test -p ml --test sp20_phase1_4_wireup_test --features cuda
-- --ignored: 2/2 pass under the new NULL-tolerant contract.
Plan reference: docs/plans/2026-05-10-sp21-train-eval-coherence-isv-defrost.md
Tier 3 status: T3.1 ✓, T3.2 ✓, T3.3 ✓ (this commit), T3.4 withdrawn,
T3.5 cascade-pending, T3.6 withdrawn.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Reverts the forced H=30 diagnostic in aux_horizon_update_kernel.cu
(commit c78c4766c) — the experiment confirmed the data ceiling
hypothesis (aux_dir_acc dropped 0.46→0.50 with pred_tanh collapsing
to ~0, opposite of the bootstrap-failure prediction). Restores the
adaptive Pearl-A + Wiener-α floor logic.
Adds SP21 plan: train/eval coherence + ISV defrost. Catalogs 26
findings across 5 tiers from a pair-audit of MinIO-archived training
logs (xmd6b 30 epochs, d7bj7 2 epochs).
Cross-cutting principle: every threshold or bound in training
control flow must be signal-driven from an ISV slot, not hardcoded
(per feedback_isv_for_adaptive_bounds + feedback_adaptive_not_tuned).
Canonical pattern documented: ISV[CURIOSITY_PRESSURE_INDEX=346]
(SP11 Fix 39) is the reference implementation.
Meta-finding: across SP3-SP20 we've been monitoring training-rollout
metrics (Thompson-noisy) instead of val metrics (deterministic
backtest). val_PF=1.18-1.33 with WR=46-48% across 30 epochs of
xmd6b shows the policy DOES extract asymmetric-payoff alpha — we
just couldn't see it through the meter inflation.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Restore 45-action factored space via Branching DQN (Tavakoli 2018),
outputting 11 Q-values (5+3+3) instead of 45. This was reduced to 5
exposure-only actions during debugging and was never intended as permanent.
- Enable use_branching: true by default in DQNConfig and DQNHyperparameters
- Add branching paths to select_action_with_confidence and select_action_inference
- Update agent.rs select_action_factored for branching-aware selection
- Expand CountBonus to per-branch tracking with bonuses_branched()
- Add order_type + urgency distribution tracking in monitoring
- Add DQN_ORDER_ACTIONS=3, DQN_URGENCY_ACTIONS=3, DQN_TOTAL_ACTIONS=45 to CUDA header
- Fix 7 pre-existing clippy doc_markdown errors in regime_conditional.rs
- Fix pre-existing cognitive_complexity in replay_buffer_type.rs (extract helpers)
- Fix flaky GPU test OOM under parallel execution (CPU fallback + test VRAM safety)
- Delete unused flash_attention submodules (block_sparse, causal_masking, etc.)
- Add GPU hot-path guard scripts and ensemble/hyperopt adapter improvements
Tests: ml-dqn 416/0, ml 905/0, clippy 0 errors on both crates
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
In candle-core (git 671de1d), min(D) and max(D) return Result<Tensor>,
not Result<(Tensor, Tensor)>. Use flatten_all() then min(0)/max(0) for
scalar reduction instead of the tuple destructuring pattern.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Task 14: evaluate_baseline is at crates/ml/examples/evaluate_baseline.rs
(not bin/fxt/src/commands/), fix cargo check command and git add path
- Task 15: process_bar_factored takes (usize, &OHLCVBarF32, &FactoredAction)
not (f64, usize, f64, f64), fix test to use correct types
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
15-task plan covering Phase 1 (seal GPU→CPU roundtrips in training loop),
Phase 2 (vectorized GPU backtest environment for hyperopt), and Phase 3
(general GPU backtester integration).
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Three-phase plan to eliminate all GPU→CPU roundtrips from training:
- Phase 1: Seal training loop (persistent GPU epoch state, async monitoring)
- Phase 2: Vectorized CUDA backtest kernel for hyperopt evaluation
- Phase 3: General-purpose GPU backtester replacing CPU SIMD path
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Bring Branching Dueling Q-Network (Tavakoli 2018) to full Rainbow parity
with the existing GPU hotpath. 3 independent advantage heads (exposure=5,
order=3, urgency=3) decompose the 45-action space into learnable branches.
H1 - CUDA fallback: gate GpuExperienceCollector when use_branching=true
(fused kernel hardcodes NUM_ACTIONS=5, incompatible with 45 factored)
H2 - Per-branch C51 distributional: each branch outputs [batch, n_d, atoms]
log-softmax, loss = avg of D cross-entropies vs projected Bellman target
M1 - NoisyNet: MaybeNoisyLinear enum in branch heads, reset_noise/disable_noise
wired through select_action, compute_loss, and set_eval_mode
M2 - Regime-conditional IS weights: Trending=1.2, Ranging=0.8, Volatile=0.6
applied to branching loss via ADX/CUSUM features at state[40:41]
M3 - State dim alignment: align_dim_for_tensor_cores() in from_dqn_params()
for H100 HMMA dispatch (8-byte alignment)
L1 - Fill simulator: splitmix64 replaces golden ratio hash (chi-squared tested)
L2 - Hyperopt 29D: branch_hidden_dim [64,256] added to PSO search space
Config plumbing: branch_hidden_dim, v_min/v_max/num_atoms, use_distributional,
use_noisy, noisy_sigma_init all flow from DQNConfig → BranchingConfig.
10 files, +3207/-125 lines, 33 branching tests + 387 ml-dqn + 284 ml-core pass.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Plan to reduce ml crate from 91K to ~12K LOC by extracting trainers
into model sub-crates, hyperopt adapters into ml-hyperopt, and
infrastructure into existing sub-crates. ml becomes an orchestration
layer owning inference, model factory, and training pipeline.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Split ml-rl into ml-dqn and ml-ppo for 3-way parallel compilation
and better sccache hit rate. Extract TradingAction to ml-core to
enable full DQN/PPO independence.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Split the monolithic ml crate (260K lines, 55s compile) into 5 crates:
ml-core, ml-rl, ml-supervised, ml-infra, ml (facade).
15-task plan with full module inventory and import migration guide.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Replace MinIO binary distribution with GitLab Generic Package Registry.
Every code push to main auto-creates a CalVer tag (vYYYY.MM.N),
compiles, uploads binaries to GitLab packages, creates a Release
with auto-generated notes, and deploys via deployment patching.
- New CI templates: create-tag, upload-release
- Modified: compile-services/training upload to GitLab packages
- Modified: deploy-services patches FOXHUNT_RELEASE on deployments
- All 7 service initContainers fetch from GitLab (curl, deploy token)
- Training job-template binary fetch from GitLab (data stays MinIO)
- MinIO retains: sccache, training data, model checkpoints
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
1. Walk-forward windows: replaced 3 non-overlapping with sliding (50% overlap, ~5 windows).
Aggregation changed from mean-0.5*std to median-0.5*IQR for outlier robustness.
2. Composite score: tanh normalization prevents Calmar ratio scale dominance
(0.02% drawdowns → values in thousands drowning out Sharpe/Sortino).
3. Q-value overestimation: new Prometheus gauge foxhunt_training_q_overestimation_ratio,
warning log when ratio>10 or q_mean>5, adaptive tau doubles when Q-mean growth>0.5/epoch
(capped at 0.01), decays back when stable.
2742 tests pass, 0 failures.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Addresses eval/training mismatch (B1-B3), reward architecture (C1-C3),
and early stopping (C4). See design doc for full analysis.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Unify both gateways into a single gRPC-only `services/api/` with tonic-web
for browser access. Drop REST+WebSocket, keep full 6-layer auth.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Replaces monolithic compile-services with 8 per-service compile jobs,
each with dependency-aware changes: filters. Deploy job only restarts
services whose binary actually changed. Also renames service crates
from snake_case to kebab-case to match k8s deployment names.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Add live training metrics monitor CLI command (streaming & one-shot) using
the monitoring gRPC service. Update DQN tests to match post-fix defaults:
IQN disabled, CQL alpha=0.1, v_min/v_max widened, 26D search space.
- train.rs: `fxt train monitor [--once] [--model X] [--interval N]`
- Rewrite gradient collapse test for BF16 mixed precision awareness
- Update inference test config to match trainer defaults (IQN off, CQL on)
- Update production smoke test for 26D parameter space
- Add dqn_action_collapse_fix_test.rs verifying all 6 root cause fixes
- Add planning docs for monitoring service and epoch financial metrics
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Helm values (controller + server on platform node, MinIO artifact repo)
- WorkflowTemplate: parameterized 5-step DAG (fetch→hyperopt→train→eval→upload)
- Nginx proxy for argo.fxhnt.ai → Argo Server :2746
- DNS A record for argo.fxhnt.ai
- MinIO bucket foxhunt-training-results for Argo artifacts
- Kustomization for kubectl apply -k
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Full GitLab CI → Argo migration: Argo Events webhook trigger,
per-service compile WorkflowTemplates, selective deploy, test
workspace, training compile, and IaC templates.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Eliminate foxhunt-binaries and training-binaries PVCs — services and
training jobs now fetch binaries from MinIO via rclone initContainers.
This removes L40S GPU autoscale-for-PVC-writes, removes RWO node
affinity constraints, and requires zero image rebuilds.
Changes:
- .gitlab-ci.yml: replace ~115 lines of binary-writer pod logic with
aws s3 cp to MinIO (~25 lines)
- 8 service YAMLs + 2 GPU overlays: add rclone initContainer + emptyDir
- Training job template: MinIO rclone fetch replaces PVC copy
- Delete foxhunt-binaries-pvc.yaml and training-binaries-pvc.yaml
- Add s3.fxhnt.ai DNS record (Terraform) and nginx proxy block
- Replace pod-writer-deploy skill with MinIO-based deploy skill
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
17-task plan for replacing hardcoded mock data in fxt watch
with real gRPC streaming data from the API Gateway.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Streaming-first architecture: 11 new poll-to-stream gateway adapters,
channel-based DataFetcher in the TUI, auto-reconnect with backoff.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
fxt 2.0 overhaul: consolidated proto/, absorbed monitoring into API Gateway,
new 15-command CLI with gRPC, MCP server mode, TUI cockpit framework.
159 files changed, net -11,835 lines.
4-phase plan from deep audit: critical bug fixes (weight_decay, IQN PER,
q_value_std, 54/51 dim mismatch), train/eval parity (TradeExecutor,
checkpoint validation), SOTA improvements (differential Sharpe, PQN
LayerNorm, temporal PER, dormant neuron resets), and search space cleanup.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Integrate all PPO improvement modules into the core training paths:
- Symlog value predictions in compute_value_loss (MLP + LSTM)
- Adaptive entropy auto-tuning replaces fixed entropy_coeff
- Percentile P5/P95 advantage scaling for heavy-tailed returns
- DAPO clip_epsilon_high wired in all 7 PPOConfig construction sites
- Shape mismatch fix in adaptive_entropy (unsqueeze scalar)
2726 tests pass, 0 clippy errors.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Clean rewrite into full-scale operations platform.
Single proto/ root, monitoring_service removed,
CLI + MCP + cockpit TUI with purple/cyan theme.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
A+B approach: smooth penalties + position limits + CUDA cleanup +
search space reduction (45D→25D) + TPE optimizer replacing PSO.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Dynamic dtype detection (Ampere+ → BF16, else F32), zero casts in
training hot path, cast only at data ingestion and loss scalar.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
