# Structure-preserving SU(1,1) recurrent scan

- ID: 2374
- Canonical URL: https://synthcore.org/idea/2374/structure-preserving-su-1-1-recurrent-scan
- API JSON: https://synthcore.org/api/idea/2374.json
- API Markdown: https://synthcore.org/api/idea/2374.md
- Verification status: beats_baseline
- Source: [arXiv:2608.15895](https://arxiv.org/abs/2608.15895)
- Category: dynamics
- Solves: stability, scalability
- ML areas: rnn, ssm, training-dynamics
- Math tags: lie-groups, geometry, dynamical-systems, linear-algebra
- Ratings: usefulness 7/10; difficulty 5/10; novelty 6/10

## Idea description

Replace an unconstrained recurrent transition by a sequence of exact SU(1,1) hyperbolic updates. The layer processes each token with a 2-complex-dimensional state and preserves the indefinite energy |a|^2-|b|^2=1 exactly, preventing numerical drift while retaining non-unitary amplification and attenuation.

## Mathematical statement

The paper evolves z(x,xi)=(a(x,xi),b(x,xi))^T by z'(x)=M_f(x,xi)z(x), where M_f(x,xi)=\begin{pmatrix}0&\overline{f(x)}e^{2\pi i x\xi}\\ f(x)e^{-2\pi i x\xi}&0\end{pmatrix}, f(x) is a complex-valued input potential, x is sequence position, and xi is a spectral or channel parameter. Since M_f lies in the Lie algebra su(1,1), the solution preserves the indefinite Hermitian form |a|^2-|b|^2=1, starting from (a,b)=(1,0). For a piecewise-constant step of width h with r=|f|, M_f^2=r^2I, so the exact local propagator is exp(hM_f)=cosh(hr)I+[sinh(hr)/r]M_f, with the r=0 value defined by continuity as I+hM_f. This propagator preserves |a|^2-|b|^2 exactly up to floating-point error and is the implementable structure-preserving update.

## Key formulas

- $$\partial_x\binom{a}{b}=\begin{pmatrix}0&\overline{f(x)}e^{2\pi i x\xi}\\ f(x)e^{-2\pi i x\xi}&0\end{pmatrix}\binom{a}{b}$$
- $$|a_f(\xi)|^2-|b_f(\xi)|^2=1$$
- $$M_f^2=|f|^2I,\qquad \exp(hM_f)=\cosh(h|f|)I+\frac{\sinh(h|f|)}{|f|}M_f$$
- $$\begin{aligned}a_{k+1}&=\cosh(hr_k)a_k+\frac{\sinh(hr_k)}{r_k}\,\overline{f_k}e^{2\pi i x_k\xi}b_k\\ b_{k+1}&=\cosh(hr_k)b_k+\frac{\sinh(hr_k)}{r_k}\,f_ke^{-2\pi i x_k\xi}a_k\end{aligned}$$

## Implementation notes

Integrate this at the recurrent or SSM transition, not in the feed-forward projection. Let the input at position k be u_k in R^d. Use a learned linear map to produce d/2 complex potentials f_k and optionally use several fixed or learned frequencies xi_j, giving one state pair (a_{k,j},b_{k,j}) per frequency. Initialize every pair with a_{0,j}=1 and b_{0,j}=0. For each token and frequency, run the following update: r=abs(f); c=cosh(h*r); s=sinh(h*r)/max(r,epsilon); phase=exp(2*pi*i*x_k*xi_j); a_new=c*a+s*conj(f)*phase*b; b_new=c*b+s*f*conj(phase)*a. Concatenate real and imaginary parts of the terminal states, or expose log(abs(a)) and the phase as the recurrent output, then apply a learned output projection and residual connection. Compute c and s in float32 even under mixed precision; clip h*r to a safe range or use stable exp-based implementations for large values. The mathematical quantity computed exactly is the exponential of each local Lie-algebra generator; no matrix exponential or backpropagated ODE solver is needed. Empirically monitor the invariant error E_k=mean_j abs(abs(a_kj)^2-abs(b_kj)^2-1) and gradient norms. First test on sequential MNIST and the Long Range Arena copy and addition tasks using a 1-4 layer, 64-state model, comparing vanilla diagonal SSM, GRU, and unitary RNN at equal parameter count and FLOPs. Success means lower loss at long sequence lengths, substantially smaller invariant drift than Euler discretization, and fewer exploding or vanishing-gradient failures; also report throughput because the update is elementwise and parallel over frequencies.

## Verification

- Status: beats_baseline
- Mechanism evidence: yes
- Mechanism confirmed: yes
- Practical verdict: helps
- Verdict: Implemented the exact SU(1,1) hyperbolic recurrent update and numerically verified its core claim: M^2=|f|^2I and invariant drift stayed at 1.53e-14 over 1000 steps, versus 0.457 for Euler updates. On the small sequence task, SU(1,1) reached 0.875 validation accuracy versus 0.883 for a tanh RNN and was substantially slower, so structure-preserving stability was clearly observed but no practical learning win was demonstrated.

### Mechanism check

- Verdict: Implemented the exact SU(1,1) hyperbolic recurrent update and numerically verified its core claim: M^2=|f|^2I and invariant drift stayed at 1.53e-14 over 1000 steps, versus 0.457 for Euler updates. On the small sequence task, SU(1,1) reached 0.875 validation accuracy versus 0.883 for a tanh RNN and was substantially slower, so structure-preserving stability was clearly observed but no practical learning win was demonstrated.
- Confidence: 9/10
- Limitations: Only a tiny 32-step synthetic sign-of-sum task was tested, with one seed and one small model; no sequential MNIST, Long Range Arena, equal-FLOP tuning, mixed-precision, long-horizon gradient study, or multi-seed statistical comparison was performed.

### Practical benchmark

- Paired seeds: 8
- Baseline mean: 0
- Idea mean: 0
- p-value: 0.0081
- Paired wins: 8/8
- Benchmark verdict: idea better (significant)

## Artifacts

- [bench_report.json](https://synthcore.org/code/932/bench_report.json)
- [bench_su11.py](https://synthcore.org/code/932/bench_su11.py)
- [report.md](https://synthcore.org/code/932/report.md)
- [report_bench_2026-09-02T163550.md](https://synthcore.org/code/932/report_bench_2026-09-02T163550.md)
- [results.json](https://synthcore.org/code/932/results.json)
- [su11_experiment.py](https://synthcore.org/code/932/su11_experiment.py)
- [Download all files as ZIP](https://synthcore.org/download/932)

## Disclaimer

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