{
 "artifacts": null,
 "category": "dynamics",
 "description": "Replace the recurrent transition or state-space mixer with a reversible transport followed by complementary relaxation of symmetric and antisymmetric feature components. The construction preserves a weighted energy and damps both parity sectors, giving bounded long-horizon powers without requiring the learned transition matrix itself to be symmetric. A numerical-range ellipse can be used as a cheap training-time certificate against transient growth.",
 "formulas_latex": [
  "$$P_{+}=\\frac{I+R}{2},\\qquad P_{-}=\\frac{I-R}{2},\\qquad C=I-s_{+}P_{+}-s_{-}P_{-},\\qquad s_{+}+s_{-}=2,\\quad 0\u003cs_{-}\u003c2.$$",
  "$$U^{\\mathsf T}HU=H,\\qquad RUR=U^{-1},\\qquad A=UC.$$",
  "$$W(B)=\\{x^{*}Bx:\\|x\\|_{2}=1\\}\\subseteq E\\quad\\Longrightarrow\\quad \\|p(B)\\|_{2}\\le(1+\\sqrt{2})\\max_{z\\in E}|p(z)|.$$",
  "$$\\|x\\|_{H}^{2}=x^{\\mathsf T}Hx,\\qquad \\|A^{n}\\|_{H}=\\sup_{x\\ne0}\\frac{\\|A^{n}x\\|_{H}}{\\|x\\|_{H}}.$$"
 ],
 "id": 2877,
 "implementation": "Integrate this as the recurrent transition in a small RNN, linear state-space model, or sequence mixer. Store the hidden state with shape [batch, positions, directions, features], and arrange directions in opposite pairs. Implement R as a fixed channel permutation. For every state f, compute f_plus = (f + R(f))/2 and f_minus = (f - R(f))/2, then apply complementary relaxation: g = f - s_plus*f_plus - s_minus*f_minus. Parameterize s_minus = 2*sigmoid(a) and set s_plus = 2 - s_minus, either globally or per feature. Apply U after relaxation. An exact low-cost U can be a fixed cyclic shift over positions followed by learned orthogonal feature mixing; parameterize the mixer with Householder reflections or obtain its orthogonal factor using QR retraction after each optimizer step. The recurrent update is f_next = U(g) + B x_t, followed by a nonlinear readout outside the recurrent transition. For diagnostics, estimate ||A^k|| by automatic-differentiation Jacobian-vector products and 10–20 power iterations. Approximate the numerical range by sampling normalized random vectors v_j and recording v_j^* A v_j; fit an empirical ellipse and penalize samples outside the unit disk or with ellipse radius above one. The exact mathematical ingredients are the parity split, complementary rates, orthogonality, and reversibility; ellipse parameters are estimated empirically. First test on a 1,000-step copy task and Tiny Shakespeare using a 2-layer, 128-dimensional model. Compare vanilla tanh RNN, orthogonal RNN, and this block at matched parameter count and FLOPs. Record gradient norm versus sequence length, maximum hidden-state amplification, validation loss, and time-to-target loss. Success means bounded amplification and non-diverging gradients at lengths where the vanilla RNN fails, with equal or lower perplexity and less than 10% extra compute.",
 "math_summary": "Let f_t in R^(N x d x q) be populations at N positions, with q paired velocity or channel directions and feature dimension d. Let R be the involution that swaps every direction with its opposite, so R^2 = I. Define parity projectors P_+ = (I+R)/2 and P_- = (I-R)/2. The two-relaxation collision is C = I - s_+ P_+ - s_- P_-, where s_+ = 2 - s_- and 0 \u003c s_- \u003c 2; therefore both parity eigenvalues 1-s_+ and 1-s_- lie in (-1,1). Let U be the transport or mixing operator and H be a positive-definite equilibrium metric. Require metric unitarity U^T H U = H and reversibility R U R = U^(-1). The recurrent transition is A = U C. The paper reduces the dynamics to a two-step recurrence generated by a contraction and bounds the associated companion operator using its numerical range W(B) = {x* B x : ||x||_2 = 1}. If W(B) is contained in an ellipse E, the Crouzeix–Palencia inequality gives ||p(B)||_2 \u003c= (1+sqrt(2)) max_{z in E} |p(z)| for every polynomial p. In the simplest neural implementation H = I; a diagonal H can model feature-dependent scales.",
 "math_tags": [
  "dynamical-systems",
  "linear-algebra",
  "geometry",
  "numerical-analysis",
  "spectral-theory"
 ],
 "ml_areas": [
  "rnn",
  "ssm",
  "training-dynamics",
  "initialization"
 ],
 "paper": {
  "arxiv_id": "2608.30253",
  "arxiv_url": "https://arxiv.org/abs/2608.30253",
  "summary_what_math_gives_to_ml": "The paper provides a transferable recipe for constructing recurrent linear operators whose powers remain uniformly bounded despite non-normal ordering of transport and relaxation. Its key ingredients are an equilibrium-metric unitary transport, velocity-reversal symmetry, and paired even/odd relaxation rates satisfying s_+ + s_- = 2 with both rates in (0,2). This suggests a recurrent or state-space neural block in which feature channels are split into reversible pairs, mixed by an orthogonal or metric-unitary operator, and damped by complementary relaxation. The numerical-range ellipse and Crouzeix–Palencia argument also provide a practical spectral diagnostic for detecting transient amplification before expensive training.",
  "title": "Mesh-Uniform Power Stability of Two-Relaxation-Time Vector Lattice Boltzmann Schemes with Reversible Boundaries",
  "year": "2026"
 },
 "ratings": {
  "difficulty": 5,
  "novelty": 5,
  "usefulness": 6
 },
 "solves": [
  "stability",
  "generalization",
  "scalability"
 ],
 "title": "Reversible Two-Relaxation Recurrent Block",
 "url": "https://synthcore.org/idea/2877/reversible-two-relaxation-recurrent-block",
 "verification": {
  "peer_reviewed": false,
  "status": "unverified",
  "status_label": "Unverified",
  "verdict_source": "deterministic test code (paired-seed permutation statistics)"
 }
}