✗ Failed on benchmark 2026

Limiter-Smoothing Bifurcation Guard

Implementation & benchmark of arXiv:2607.18894 — Bifurcation Analysis of Sub-Synchronous Oscillations Related to Grid-Forming Converter Inner Controllers

Usefulness7/10

Difficulty5/10

Novelty6/10

Regularization Stability Accuracy Optimizer Training dynamics Regularization Dynamical systems Bifurcations Control theory Numerical analysis

Source paper: Bifurcation Analysis of Sub-Synchronous Oscillations Related to Grid-Forming Converter Inner Controllers arXiv:2607.18894 ⓘ · analyzed Aug 30, 2026

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 Not peer-reviewed.

Idea description

Use the paper's finding that smooth approximations of a circular current limiter can generate spurious Hopf bifurcations to audit smooth bounded operations in optimizers and networks. Compare exact projection with a differentiable surrogate, continue both dynamics in clipping threshold or step size, and reject a surrogate if it introduces a unit-circle crossing absent from the exact map.

Formulas

$$P_R(v)=v\min\left(1,\frac{R}{\|v\|_2}\right),\quad S_{R,\epsilon}(v)\approx P_R(v).$$

$$z_{k+1}=F_P(z_k,p)\quad\text{or}\quad z_{k+1}=F_S(z_k,p),\quad J_{P/S}=\frac{\partial F_{P/S}}{\partial z}.$$

$$\rho(J_S(p_c))=1\text{ with }\lambda_{1,2}=e^{\pm i\theta},\quad \rho(J_P(p))<1\text{ for the same }p.$$

$$\|P_R(v)\|_2\le R,\quad \|S_{R,\epsilon}(v)\|_2\text{ can have a different local gain.}$$

Mathematical statement

Let P_R(v) be exact projection of an update vector v onto a radius-R Euclidean ball, and let S_(R,epsilon)(v) be a differentiable approximation. The two optimizer maps are F_P and F_S, with Jacobians J_P and J_S. A smoothing-induced spurious instability exists if continuation finds a parameter p_c satisfying rho(J_S(p_c))=1 with a complex eigenvalue pair, while the exact projected map remains stable, rho(J_P(p))<1, over the same interval. The mechanism is falsifiable by comparing bifurcation diagrams, spectral crossings, and perturbed trajectories for exact and smoothed clipping.

Implementation notes

Integration point: replace gradient or optimizer-state clipping in a small neural network with two alternatives: exact Euclidean projection and a smooth radial approximation such as S_(R,epsilon)(v)=v/sqrt(1+(||v||_2/R)^2). Keep clipping radius R, smoothing width epsilon, learning rate eta, and momentum beta explicit. Pseudocode: define one-step maps F_P and F_S including model and optimizer states; for a grid of eta, R, and epsilon, solve or approximate a stationary state; compute Jacobian-vector products for both maps; use Arnoldi iteration to estimate leading eigenvalues; continue stationary states in eta or R with predictor-corrector; record the first complex crossing where absolute value(lambda)=1; then run trajectories from identical small perturbations. The paper supplies the warning that smoothing can create artificial Hopf points and the comparison methodology. The implementation estimates Jacobians, stationary states, and stochastic-gradient effects. First experiment: train a two-layer MLP on MNIST with momentum SGD, sweep eta near the largest stable value, and compare exact versus smooth clipping. Prediction: an interval can appear where rho(J_S)>1 and loss oscillates persistently while rho(J_P)<1 and perturbations decay. As epsilon decreases, the smooth crossing should move toward the exact-map boundary or disappear.

Verification

Failed on benchmark

Stage 1 · Toy mechanism gate: Passed ✓

Stage 2 · Mechanism transferred to benchmark: Not reproduced ✗

Stage 2 · Practical benchmark result: Inconclusive

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 ran a two-dimensional exact-projection versus smooth-radial-limiter audit. Exact and smooth crossing predictions matched numerical sweeps, finite-difference Jacobian errors were below 2.5e-10, and an intermediate gain produced decay for exact projection but growth for the smooth surrogate. This supports the mechanism in the toy system.

Agent confidence: 8/10

Baseline: Exact boundary 3.311644 for K=1,2,4,8; final perturbation norm 1.2.
Idea: Smooth boundaries 2.44, 1.725341, 1.22, 0.862670; K=4 trajectory final norm 3.67065.

Limitations: Only a deterministic two-dimensional toy was tested; no MNIST/MLP benchmark, epsilon sweep, stochastic optimizer, Arnoldi iteration, or predictor-corrector continuation was implemented.

How to run: python3 limiter_guard.py

Stage 2 — Benchmark vs. tuned baseline (arXiv:2607.18894) Sep 2, 2026

Ahead of the baseline (−18.1%) but not statistically significant (p=0.2949, wins 5 / 8 paired seeds).

Benchmark: Dynamics forecasting
Model: rnn_small
Paired seeds: 8
Baseline mean: 0.0012
Idea mean: 0.001
Effect (Δ): -0.0002 (−18.1%; negative = idea better)
Wins: 5 / 8 paired seeds
p-value: 0.2949 (permutation test, 20 000 shuffles)
Smallest detectable effect: n/a
Mechanism: Not confirmed ✗
Practical effect: Inconclusive
Baseline tuning: swept over 3 configs

Limitations:

Only the built-in dynamics track was tested; no Arnoldi Jacobian continuation, stationary-state bifurcation scan, epsilon sweep, or persistent-oscillation trajectory analysis was performed. The benchmark used three shared learning rates and one clipping radius, 12 epochs, and 400/160 train/test samples.

How to run: python3 stage2_limiter_guard.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: 9/10

Implemented exact global gradient projection versus the differentiable smooth radial surrogate in the structurally matched dynamics track using identical rnn_small systems, Adam settings, data, epochs, and paired seeds. The idea's best mean test MSE was 0.0010236 versus baseline 0.0012496, but the paired delta was -0.0002260 with permutation p=0.2949, hence no significant win. The trained-model signature did not confirm the predicted larger smooth applied norm: exact p95 was 0.0500000 while smooth p95 was 0.0499769.

Baseline: best_cfg lr=0.003, radius=0.05; full 8-seed mean MSE=0.0012495772, std=0.0005645325
Idea: best_cfg lr=0.006, radius=0.05; full 8-seed mean MSE=0.0010236013, std=0.0003525166; paired delta=-0.0002259759; p=0.2949; wins=5/8

Benchmark result