cascade-fullctx-spectral-v12

A differentiated NumPy/SciPy full-context generator derived from the public custom-fullctx-v4 design.

Design

• Emits only 4096-point series, matching the 128-patch training/eval geometry and minimizing the one non-target context patch paid per series.
• Samples trend as total end-to-end excursion, so trend strength is independent of series length.
• Mixes fourteen families: trend/seasonality, structural regimes, multiplicative series, AR/integrated/nonlinear dynamics, smooth spectral GP, power-law long memory, physical sensors, seasonal counts, intermittent demand, and outliers.
• Replaces the predecessor's slow 48-pass random-Fourier GP with one batched inverse FFT and adds persistent/anti-persistent spectral paths.
• Adds a small regime-switching mean-reverting family with bounded clustered volatility, heavy-tailed innovations, transient shocks, and seasonal means.
• Executes the OU recurrence with SciPy's compiled linear filter instead of a 4095-step Python loop.
• Executes AR(1) and AR(2) recurrences with compiled SciPy filters; local microbenchmarks were 4–12× faster than scanning time in Python.
• Adds batched cadence-seasonal Poisson and gamma-mixed Poisson counts with signed trends and decaying bursts, preserving positive integer structure.
• Adds weekday/weekend interactions to a minority of count rows and piecewise spectral slopes to long-memory rows, covering calendar effects and scale-dependent roughness without another FFT.
• Gives 40% of trend-seasonal rows a low-innovation mode, teaching sharp, stable periodic reconstruction while retaining noisy seasonal coverage.
• Anchors most mixture mass on the five-seed-tested full-context core (trend/seasonal, regimes, multiplicative, AR2, and integrated paths), while retaining each newer prior at a conservative share.
• Uses published TempoPFN ablations to strengthen OU/SDE-like dynamics, spectral/long-memory paths, and transient events without letting one family dominate the corpus.
• Adds slowly modulated amplitude and phase to a minority of seasonal components; stationary cycles remain the majority. Multiplicative paths use the same full cadence bank instead of a four-period subset.
• Extends structural/event coverage with piecewise-affine regime trends, decaying shock recovery, event plateaus, and genuine held-constant runs.
• Applies low-rate reversal, censoring, quantization, and sample-and-hold artifacts to bridge clean priors to real measurement pipelines.
• Extends seasonality through 365/672/730-step cycles and adds a small generic physical-sensor family (smooth, bounded, pressure-like, and skewed-positive) without adopting the competitor's private-pool-shaped weather weighting.
• Generates lazy 2048-row random-family chunks, keeping every stream prefix mixed while amortizing Python dispatch. Local profiling found this about 6% faster than 1024 rows; 4096 rows regressed slightly.
• Evaluates optional seasonal components only for active rows while preserving the fixed RNG draw sequence, and caches the fixed cadence sine/cosine basis, reducing trigonometric work without narrowing the prior.
• Draws jump, shock, and heavy-tail values only where those sparse branches are active instead of allocating dense arrays whose values are mostly discarded.

On this VPS, an 8192-series benchmark improved from a v11 pre-optimization median of 9.40M points/s to 10.72M points/s (five-run median), about 14%. The generator is now well above the mainnet contract's 3.7M reference throughput in isolation; end-to-end token completion also includes model training and stream handoff.

A controlled 120-second parameter screen compared the baseline mixture with seasonal-, spectral-, and dynamics-heavy variants under the same model, pool, budget, and seeds. Dynamics-heavy won all three validation seeds, reducing mean local synthetic-pool geomean from 0.19097 to 0.18431 (3.5%; lower is better).

v12 then incorporates the live j-test/2_1_v5 king's strongest transferable signal: more mass on learnable long-context trend, regime, AR(2), integrated, and smooth spectral structure, with sparse/choppy families kept small. It uses a conservative blend rather than copying the king's exact weights because this generator adds OU, long-memory, physical, and count priors absent from the king. The king's older Python recurrences and 48-pass RFF GP are not copied. This remains a quality hypothesis until production-faithful GPU A/B testing.

Local training result

The preceding v10 corpus was trained under the mainnet chain.toml contract on an A100 for the full 3-hour wall. It scored 0.13679 on the 64-window local synthetic smoke pool (lower is better), improving from 0.15429 at the 30-minute heat budget. It reached 55% of the contract token budget, so v11 targets generator throughput rather than changing family weights from a single non-private-pool result. These scores are directional and are not live validator verdicts.

Validate

python -m cascade.miner.cli verify ./cascade-v2 --chain-toml chain.toml

Contract validity and CPU throughput do not establish forecasting quality. Run a production-faithful GPU A/B score against the current king before deploying this candidate.