{
 "artifacts": null,
 "category": "dynamics",
 "description": "Partition a neural network into interacting modules and constrain the product of their local finite-region gains and coupling strengths so that the resulting gain matrix has spectral radius below one. This transfers the paper's small-gain-like mechanism and gives a quantitative large-signal boundary: instability or exploding activations should emerge as the spectral radius approaches one, while a weighted Lyapunov function should contract below that boundary.",
 "formulas_latex": [
  "$$z_i^+\\leq a_i z_i+b_i\\sum_{j=1}^{n}|c_{ij}|r_j z_j,\\qquad z_i=\\|x_i-\\tilde{x}_i\\|,$$",
  "$$G_{ii}=a_i,\\qquad G_{ij}=b_i|c_{ij}|r_j\\ (i\\neq j),\\qquad \\rho(G)\u003c1,$$",
  "$$p^\\top G\\leq\\kappa p^\\top,\\quad p\\succ0,\\quad 0\u003c\\kappa\u003c1,\\qquad V(z)=p^\\top z,$$",
  "$$\\mathcal E_r=\\{\\delta x:\\delta x^\\top P\\delta x\\leq r^2\\},\\qquad \\sup_{\\delta x\\in\\mathcal E_r}\\frac{V_2(\\delta x^+)}{V_2(\\delta x)}\\leq\\kappa^2\u003c1.$$"
 ],
 "id": 2838,
 "implementation": "Use this at the boundary between residual blocks, recurrent modules, or MoE experts rather than on every individual scalar neuron. Write the model as x_i^+=f_i(x_i,u_i) and expose explicit couplings u_i=sum_j c_ij h_j(x_j); for a transformer, modules can be attention, MLP, normalization, and residual streams, while for an RNN they can be state-update and feedback branches. At each training checkpoint, estimate a_i and b_i by power iteration or minibatch Jacobian-vector products over a prescribed activation region: estimate a_i approximately as the maximum of ||partial f_i/partial x_i||_2 and b_i as the maximum of ||partial f_i/partial u_i||_2 over random and high-norm activation samples. Estimate r_j similarly for h_j, construct G, and compute its spectral radius by eigendecomposition. Add the penalty lambda max(0,rho(G)-(1-epsilon))^2 to the task loss, or rescale coupling weights whenever rho(G)\u003e1-epsilon. Pseudocode: compute local gains; form G; calculate rho(G); backpropagate task loss plus gain penalty; optionally project edge weights by c_ij \u003c- c_ij(1-epsilon)/rho(G) when the boundary is violated. Separately fit P positive definite from empirical covariance or solve a discrete Lyapunov inequality for a linearized module graph, then test the ellipsoid contraction ratio. The paper contributes the small-gain assembly and invariant-region logic; Jacobian norms, gain estimation, and projection are engineering adaptations. The first cheap experiment should use a 4-block residual MLP on CIFAR-10 or a delayed-copy sequence task, comparing unconstrained residual training with the spectral-radius penalty at equal parameter count. Sweep a global coupling multiplier alpha and predict a sharp transition near alpha_c=1/rho(G_0), where G_0 is the unscaled measured gain matrix: activation perturbations and long-horizon Jacobian norms should decay for alpha\u003calpha_c and grow for alpha\u003ealpha_c, with the observed boundary within 20% of prediction. Also test whether the estimated ellipsoid contraction ratio remains below one for perturbations inside E_r and crosses one as r increases beyond the certified region.",
 "math_summary": "Let module i in {1,...,n} have state or activation x_i and input u_i, with update x_i^+=f_i(x_i,u_i). Assume the finite-region incremental bound ||f_i(x_i,u_i)-f_i(tilde{x}_i,tilde{u}_i)|| \u003c= a_i||x_i-tilde{x}_i||+b_i||u_i-tilde{u}_i||, where a_i,b_i\u003e=0 are estimated local state and input gains. Let the coupling be u_i=sum_j c_ij h_j(x_j), where c_ij is an edge weight or operator norm and h_j has gain r_j. Define G_ii=a_i and G_ij=b_i|c_ij|r_j for i != j. The interconnected perturbation magnitudes satisfy z^+\u003c=Gz, with z_i=||x_i-tilde{x}_i||. If rho(G)\u003c1, where rho is the spectral radius, there exists p\u003e0 and kappa\u003c1 such that p^T G\u003c=kappa p^T. The weighted Lyapunov function V(z)=p^Tz then contracts. A quadratic empirical certificate uses V_2(delta x)=delta x^T P delta x, P positive definite, and an invariant ellipsoid E_r={delta x:delta x^TPdelta x\u003c=r^2} when the worst-case ratio V_2(delta x^+)/V_2(delta x) is at most kappa^2\u003c1 throughout E_r.",
 "math_tags": [
  "control-theory",
  "dynamical-systems",
  "linear-algebra",
  "optimization"
 ],
 "ml_areas": [
  "optimizer",
  "training-dynamics",
  "rnn",
  "transformer",
  "moe-routing"
 ],
 "paper": {
  "arxiv_id": "2608.29570",
  "arxiv_url": "https://arxiv.org/abs/2608.29570",
  "summary_what_math_gives_to_ml": "The paper provides a small-gain-like construction for large-signal stability of interconnected nonlinear subsystems: individual dissipation or gain bounds are assembled into a global stability certificate. Its most transferable asset is a computable interaction-gain matrix whose spectral radius predicts whether coupled modules remain stable, together with a Lyapunov function and ellipsoidal forward-invariant region. A neural implementation should treat residual blocks, recurrent cells, or mixture-of-experts routes as interconnected subsystems, estimate their finite-region gains, and constrain the interaction matrix below the small-gain boundary.",
  "title": "A Small-Gain-Like Framework for Large-Signal Stability Evaluation of Multi-Converter Systems",
  "year": "2026"
 },
 "ratings": {
  "difficulty": 5,
  "novelty": 6,
  "usefulness": 7
 },
 "solves": [
  "stability",
  "accuracy"
 ],
 "title": "Small-Gain Constrained Neural Modules",
 "url": "https://synthcore.org/idea/2838/small-gain-constrained-neural-modules",
 "verification": {
  "peer_reviewed": false,
  "status": "unverified",
  "status_label": "Unverified",
  "verdict_source": "deterministic test code (paired-seed permutation statistics)"
 }
}