△ Mechanism confirmed, baseline not beaten 2026

FDT-Constrained Conservative Neural Flow

Implementation & benchmark of arXiv:2607.22527 — Effective field theories of nonlinear fluctuating hydrodynamics in one dimension

Usefulness8/10

Difficulty6/10

Novelty7/10

Dynamics Stability Generalization Accuracy Ssm Rnn Training dynamics Dynamical systems Stochastic processes Control theory Statistical mechanics Pde

Source paper: Effective field theories of nonlinear fluctuating hydrodynamics in one dimension arXiv:2607.22527 ⓘ · analyzed Aug 30, 2026

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 Not peer-reviewed.

Idea description

Replace an unconstrained residual or state-space update by a discrete conservative stochastic balance law. The neural network learns nonlinear mode-coupling fluxes, while the dissipative operator and injected noise are tied by a fluctuation-dissipation relation so that the model has a controlled stationary distribution rather than unconstrained activation drift.

Formulas

$$\partial_t\phi_a(x,t)+\partial_x\!\left(v_a\phi_a+\sum_{b,c}G^a_{bc}\phi_b\phi_c-\sum_bD_{ab}\partial_x\phi_b+\xi_a\right)=0.$$

$$z_i^{t+1}=z_i^t-\frac{\eta}{h}\left(F_{i+1/2}(z^t)-F_{i-1/2}(z^t)\right)+\frac{\eta}{h^2}D\left(z_{i+1}^t-2z_i^t+z_{i-1}^t\right)+\frac{\sqrt{\eta}}{h}(\varepsilon_{i-1/2}^t-\varepsilon_{i+1/2}^t),$$

$$F_{i+1/2,a}(z)=v_a\bar z_{i+1/2,a}+\sum_{b,c}G^a_{bc}\bar z_{i+1/2,b}\bar z_{i+1/2,c}+f_{\theta,a}(z_i,z_{i+1}),\qquad \varepsilon_{i+1/2}^t\sim\mathcal N(0,Q),\quad Q=D_sC+CD_s^\top.$$

$$\eta\le\eta_c\equiv\frac{h^2}{2\lambda_{\max}(D_s)},\qquad \sum_i z_i^{t+1}=\sum_i z_i^t.$$

Mathematical statement

The paper's hydrodynamic mechanism is the conservative stochastic equation $\partial_t\phi_a+\partial_x(\mathcal J_a^{\rm R}+\mathcal J_a^{\rm D}+\xi_a)=0$, with reversible nonlinear flux $\mathcal J_a^{\rm R}=v_a\phi_a+\sum_{b,c}G^a_{bc}\phi_b\phi_c$ and dissipative flux $\mathcal J_a^{\rm D}=-\sum_bD_{ab}\partial_x\phi_b$. Here $\phi_a(x,t)$ is mode $a$, $v_a$ is its propagation velocity, $G^a_{bc}$ are mode-coupling coefficients, $D$ is a diffusion matrix, and $\xi$ is a random current. For a neural implementation, let $z_i\in\mathbb R^q$ be the state at grid position or sequence index $i$, let $C\succ0$ be the target local-equilibrium covariance, and let $D_s=(D+D^\top)/2\succeq0$. Use current-noise covariance $Q=2D_sC$ when $D_sC$ is symmetric, or the general symmetric choice $Q=D_sC+CD_s^\top$. The discrete Laplacian contribution is stable when the symmetric part of $D$ is positive semidefinite; for an explicit Euler step with grid spacing $h$, the linear diffusion mode has eigenvalues $-4\lambda_k(D_s)\sin^2(\pi k/n)/h^2$, giving the sufficient step-size bound $\eta\le h^2/[2\lambda_{\max}(D_s)]$.

Implementation notes

1. Integration point: use this layer as a replacement for a residual block, recurrent state update, or latent spatial layer. Arrange the hidden state as $z\in\mathbb R^{n\times q}$, where $n$ is the sequence or grid length and $q$ is the number of interacting modes. Compute interface states $\bar z_{i+1/2}=(z_i+z_{i+1})/2$, evaluate a small MLP $f_\theta(z_i,z_{i+1})$ for learned reversible flux corrections, and keep every update in flux-difference form. Parameterize $D_s=LL^\top$ and set $D=D_s+A$ with skew-symmetric $A$, ensuring nonnegative dissipative eigenvalues. 2. Pseudocode: initialize $z$; compute all interface fluxes $F$; sample independent Gaussian interface currents $\varepsilon\sim\mathcal N(0,Q)$; update z = z - eta/h*(F_right-F_left) + eta/h**2 * D @ laplacian(z) + sqrt(eta)/h*(eps_left-eps_right); use periodic boundaries or equal boundary fluxes. 3. Computed from the mechanism: conservative divergence structure, nonlinear quadratic mode coupling, and the FDT covariance $Q$. Estimated empirically: $C$ from latent minibatch covariance, $v,G,D$ by gradient descent, and $\lambda_{\max}(D_s)$ by power iteration. Add a penalty $\|\widehat C-C\|_F^2$ only if the target covariance is not known. 4. First cheap experiment: train a 1D latent state-space model on noisy advection-diffusion sequences, comparing an unconstrained residual SSM, deterministic conservative flux dynamics, and the FDT-constrained layer. Use $n=64$, $q=4$, and synthetic data generated from the same equation. Sweep $\eta$ across the predicted boundary $\eta_c=h^2/(2\lambda_{\max}(D_s))$. The quantitative prediction is that the constrained model preserves $\sum_i z_i$ to numerical precision, remains linearly stable below $\eta_c$, becomes unstable near or above $\eta_c$, and its long-run latent covariance approaches $C$ with error decreasing as trajectory length grows. Measure the observed divergence threshold and require agreement with $\eta_c$ within 20%. Also test whether the equilibrium power spectrum is reproduced more accurately than by unconstrained noise injection.

Verification

Mechanism confirmed, baseline not beaten

Stage 1 · Toy mechanism gate: Passed ✓

Stage 2 · Mechanism transferred to benchmark: Confirmed ✓

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 9/10

Built a conservative periodic neural-flow layer with parameterized positive-semidefinite diffusion, skew transport component, learned interface flux MLP, and FDT-tied current noise. The mechanism manifested: conservation error was 5.77e-15, the observed stability transition was 1.15 times the predicted eta_c across all diffusion values (within the requested 20%), and matched FDT covariance error was 0.013 versus 0.790 for mismatched noise. The unconstrained additive-noise baseline changed the conserved sum by 5.58 on average and had covariance error 1.682.

Agent confidence: 9/10

Baseline: Unconstrained additive noise: mean conserved-sum change 5.575; relative nonzero-mode covariance error 1.682.
Idea: Conservative FDT flow: conservation max absolute error 5.77e-15; observed stability boundary 1.15*eta_c versus predicted 1.00*eta_c; matched-FDT covariance error 0.01297 versus 0.78988 for mismatched noise.

Limitations: This is a synthetic toy verification, not a trained sequence-prediction benchmark. The neural flux MLP is instantiated and executed for shape/conservation validation but was not trained or compared on predictive loss; stability was measured on a coarse parameter grid, and only diagonal target covariance cases were tested.

How to run: python3 fdt_flow.py

Stage 2 — Benchmark vs. tuned baseline (arXiv:2607.22527) Sep 2, 2026

Significantly WORSE than the tuned baseline (n/a, p=0.0081, wins 0 / 8 paired seeds).

Benchmark: Periodic_transport
Model: local_flow_net
Paired seeds: 8
Baseline mean: 0
Idea mean: 0.0106
Effect (Δ): 0.0035 (n/a; negative = idea better)
Wins: 0 / 8 paired seeds
p-value: 0.0081 (permutation test, 20 000 shuffles)
Smallest detectable effect: n/a
Mechanism: Confirmed ✓
Practical effect: Hurts
Baseline tuning: swept over 3 configs

Limitations:

Only the custom periodic transport PDE track was tested; built-in vision, sequence, and dynamics tracks were not tested. The mechanism signature used a trained model but did not perform a full empirical stability-boundary sweep at neural-network scale.

How to run: python3 bench_fdt.py

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

Stage 2 — Benchmark Aug 30, 2026 latest Failed ✗

Agent confidence: 9/10

Built a structurally matched periodic transport PDE benchmark comparing an unconstrained residual update with a conservative flux-difference FDT update using the same encoder/readout system. The trained-model signature confirmed conservation and stable diffusion behavior, but the idea lost significantly on the independent task metric.

Baseline: Best lr=0.01, 18 epochs; mean test MSE 0.00709239, std 0.00067847.
Comparison: Paired delta idea-baseline +0.00350275 MSE; 0/8 wins; permutation p=0.0081; significant loss.
Idea: Best lr=0.01, 18 epochs; mean test MSE 0.01059514, std 0.00083073.
Mechanism signature: Predicted conservation error 0; observed trained-model error 7.15e-7; predicted eta_c 4.89336; observed amplification 0.99626; confirmed true.

Benchmark result