✗ Failed on benchmark 2026

Spectral-Band Dual-Timescale Network

Implementation & benchmark of arXiv:2607.17358 — Richards' equation as a hydrodynamic limit: Chapman--Enskog reduction of the continuum kinetic equation for unsaturated soil water

Usefulness7/10

Difficulty6/10

Novelty6/10

Architecture Stability Memory Speedup Rnn Ssm Training dynamics Inference speedup Dynamical systems Spectral theory Control theory Linear algebra Numerical analysis

Source paper: Richards' equation as a hydrodynamic limit: Chapman--Enskog reduction of the continuum kinetic equation for unsaturated soil water arXiv:2607.17358 ⓘ · analyzed Aug 30, 2026

AI-generated research hypothesis, automatically tested. Not peer-reviewed.

Idea description

Split hidden dynamics into relaxation bands when the Jacobian spectrum has a gap, evolve each band with its own timescale, and retain an explicit cross-band exchange term. This yields a principled dual-timescale RNN or SSM rather than choosing fast and slow branches heuristically.

Formulas

$$J\approx P_fJP_f+P_sJP_s,\qquad \mathrm{spec}(P_fJP_f)\subset(-\infty,-\gamma_f],\qquad \mathrm{spec}(P_sJP_s)\subset[-\gamma_s,0],\qquad \gamma_f\gg\gamma_s.$$

$$\dot h_f=\frac{1}{\mathrm{Da}}C_f(h_f)+E_{fs}h_s+B_fx,\qquad \dot h_s=\frac{1}{\mathrm{Da}}C_s(h_s)+E_{sf}h_f+B_sx.$$

$$A=\begin{bmatrix}J_f/\mathrm{Da}&E_{fs}\\E_{sf}&J_s/\mathrm{Da}\end{bmatrix},\qquad \lambda_{\max}\left(\frac{A+A^\top}{2}\right)\le0\ \Longrightarrow\ \|h(t)\|\ \text{does not grow}.$$

$$\tau_f\simeq\gamma_f^{-1},\qquad \tau_s\simeq\gamma_s^{-1},\qquad \text{reduction error increases when }\omega\tau_f\gtrsim1.$$

Mathematical statement

Let $J$ be the linearized redistribution operator and suppose its spectrum separates into fast eigenvalues at most $-\gamma_f$ and slow eigenvalues in $[-\gamma_s,0]$, with $\gamma_f\gg\gamma_s$. Let $P_f$ and $P_s$ be spectral projectors, $h_f=P_fg$, and $h_s=P_sg$. The construction applies Chapman–Enskog reduction inside each band while retaining cross-band relaxation at order one. The neural analogue uses within-band dissipative dynamics and exchange operators $E_{fs}$ and $E_{sf}$. Stability is guaranteed when the symmetric part of the full block Jacobian is negative semidefinite. The spectral gap predicts two decay timescales and a reduction breakdown when the input frequency approaches the fast relaxation rate.

Implementation notes

Implement a two-branch recurrent cell or linear/nonlinear SSM with fast state $h_f\in\mathbb{R}^{d_f}$ and slow state $h_s\in\mathbb{R}^{d_s}$. Parameterize the within-branch linear operators as $J_f=-L_f^\top L_f-\alpha_fI$ and $J_s=-L_s^\top L_s-\alpha_sI$, with trainable nonnegative $\alpha_f$ and $\alpha_s$. Add learned exchange matrices $E_{fs}$ and $E_{sf}$, then constrain the symmetric part of the complete block Jacobian by spectral normalization or projection so that $\lambda_{\max}((A+A^\top)/2)\le0$. Feed the input through separate projections $B_f$ and $B_s$. Apply the fast branch with an exact matrix exponential or a semi-implicit update, allowing larger stable macro-steps, and update the slow branch with the ordinary recurrent step. Estimate the band gap from block-Jacobian eigenvalues using a few Lanczos iterations. The paper supplies the band decomposition, within-band reduction, and order-one exchange; the neural parameterization is an implementation choice. First test on Copy Memory, sequential CIFAR-10, and synthetic signals containing two frequencies, against GRU, S4, and a single-timescale SSM. Sweep the imposed ratio $\gamma_f/\gamma_s$. Prediction: impulse responses exhibit two exponential slopes with time constants near $1/\gamma_f$ and $1/\gamma_s$; reduced dynamics becomes biased when $\omega\approx\gamma_f$, whereas retaining both bands preserves the response. Measured decay rates should agree with Jacobian eigenvalues within 20%.

Verification

Failed on benchmark

Stage 1 · Toy mechanism gate: Passed ✓

Stage 2 · Mechanism transferred to benchmark: Not reproduced ✗

Stage 2 · Practical benchmark result: Significantly worse ✗

Methodology: 8 paired seeds, permutation test (p<0.05) against a learning-rate-tuned baseline. How verification works

Stage 1 — Mechanism check agent confidence 8/10

Built and numerically verified a compact dual-timescale linear network with dissipative fast/slow branches and explicit cross-band exchange. The stability boundary matched k_crit=sqrt(gamma_f gamma_s) to 1.4e-16 relative error, measured decay-rate ratios tracked imposed spectral gaps within about 0.3%, and slow-only frequency-response error increased 3.25x near the fast relaxation rate. The dual-band representation reduced toy two-frequency reconstruction MSE by 79%, although speedup and broad task-level superiority were not tested.

Agent confidence: 8/10

Baseline: Single slow-timescale filter: MSE 0.6988269
Idea: Dual spectral-band filter with exchange: MSE 0.1467527, 79.0% lower; observed stability boundary 3.16227766 versus predicted 3.16227766; fitted rate ratios for imposed gaps 2, 5, 10, 20, 40 were 2.015, 5.019, 10.031, 20.055, and 40.105; slow-only relative transfer error was 0.117 at low frequency versus 0.381 near the fast rate

Limitations: This was a small linear NumPy toy experiment rather than a trained nonlinear RNN or SSM. GRU, S4, Copy Memory, sequential CIFAR-10, inference speed, learned spectral projection, and larger-dimensional robustness were not tested; high-frequency reduction error continued rising beyond the fast-rate transition, so no sharp error peak exactly at gamma_f was established.

How to run: python3 spectral_band_experiment.py

Stage 2 — Benchmark vs. tuned baseline (arXiv:2607.17358) Sep 2, 2026

Significantly WORSE than the tuned baseline (+29.5%, p=0.0081, wins 0 / 8 paired seeds).

Benchmark: Dynamics forecasting
Model: rnn_small
Paired seeds: 8
Baseline mean: 0.0004
Idea mean: 0.0005
Effect (Δ): 0.0001 (+29.5%; negative = idea better)
Wins: 0 / 8 paired seeds
p-value: 0.0081 (permutation test, 20 000 shuffles)
Smallest detectable effect: ±14.5%
Mechanism: Not confirmed ✗
Practical effect: Hurts
Baseline tuning: swept over 6 configs

Limitations:

Only the built-in dynamics track was tested; vision, sequence, and tabular tracks were not structurally necessary. The implementation used a compact Euler-updated recurrent cell rather than exact matrix exponentials, and the mechanism probe used a single short zero-input response per seed; its decay fit collapsed to zero and therefore did not confirm the quantitative prediction. No inference-speed or memory benchmark was performed.

How to run: python3 stage2_bench.py

Verdict computed by deterministic test code from paired-seed statistics — not by the language model.

Stage 2 — Benchmark Aug 31, 2026 latest Failed ✗

Agent confidence: 10/10

Implemented and evaluated a matched dual-timescale dissipative recurrent network with explicit cross-band exchange on the structurally appropriate dynamics track. Across 8 paired seeds, the idea was significantly worse than the tuned single-band baseline: MSE 0.0005187966 versus 0.0004004990, paired delta +0.0001182976, permutation p=0.0081. The trained-model decay signature was not confirmed because the probe response collapsed to zero, so there is no evidence that the predicted mechanism transferred beneficially to this neural benchmark.

Baseline: Tuned single-band dissipative RNN, best lr=0.006 and gamma=1.0; full 8-seed MSE mean 0.0004004990, std 0.0000735487
Delta mean: idea - baseline = +0.0001182976, approximately 29.54% worse relative to baseline
Idea: Dual-band dissipative/exchange RNN, best lr=0.006; full 8-seed MSE mean 0.0005187966, std 0.0000673149
Idea wins: 0
N pairs: 8
P value: 0.0081

Benchmark result