Abstract
Defines two notions of equality (strict vs magnitude-only) and a symbolic preorder driven by a tunable scoring functional S_beta. With beta = 0 it collapses to classical magnitude order; with beta > 0 it becomes stability-aware. ASCII-only.
🧾 Equality notions
Strict equality
(m1, a1) == (m2, a2) iff (m1 == m2) and (a1 == a2)
Magnitude equality
(m1, a1) ~= (m2, a2) iff (m1 == m2) # ignore alignment
Use magnitude equality to say “same size, different centre state.”
Illustration
Let x = (7, +1) and y = (7, -1).
- Strict:
x == y? No (alignment differs). - Magnitude:
x ~= y? Yes (both havem = 7).
SSM thus distinguishes stability even when size is identical.
📐 Ordering notions
Classical magnitude order
(m1, a1) <_m (m2, a2) iff m1 < m2
Symbolic strength (default scoring functional)
Fix beta ∈ [0, 1]. Define:
S_beta(m, a) = m * (1 - beta * (1 - a))
# beta = 0 → S_0(m,a) = m (alignment ignored)
# beta = 1 → S_1(m,a) = m * a (alignment fully weighted)
Signedness note: S_beta is signed; negative m yields negative scores unless your application restricts to m ≥ 0.
Symbolic preorder (default, total with ties resolved)
For x=(m1,a1), y=(m2,a2):
x <=_beta y iff
[ S_beta(m1,a1) < S_beta(m2,a2) ] or
[ S_beta(m1,a1) == S_beta(m2,a2) and ( a1 < a2 or (a1 == a2 and |m1| <= |m2| ) ) ]
This relation is reflexive and transitive; with the tie-break rules it is a total preorder on R × [-1, 1].
Optional absolute-magnitude variant (declare if used)
For sign-agnostic ranking:
S_beta_abs(m, a) = |m| * (1 - beta * (1 - a))
📈 Monotonicity lemmas (sketch)
- Increasing in alignment (
m ≥ 0fixed): ifaincreases, thenS_beta(m,a)increases. - Increasing in magnitude (
a ≥ 0fixed): ifmincreases, thenS_beta(m,a)increases. - Collapse consistency: on the slice
a = +1,<=_betacoincides with classical magnitude order.
(Proof sketches in the Proof Appendix.)
🧪 Worked examples
Example 1 — Alignment can outweigh size
Compare (10, -0.3) vs (9, +0.9).
- Classical:
10 > 9. - Symbolic (
beta = 1):S_1(10, -0.3) = 10 * (-0.3) = -3 S_1( 9, +0.9) = 9 * 0.9 = 8.1Hence(9, +0.9) >=_1 (10, -0.3).
Example 2 — Tuning beta switches the ranking
Compare (8, -0.5) vs (7, +0.9).
beta = 0(classical):S_0(8, -0.5) = 8,S_0(7, +0.9) = 7→(8, -0.5)ranks higher.beta = 1(fully stability-aware):S_1(8, -0.5) = -4,S_1(7, +0.9) = 6.3→(7, +0.9)ranks higher.
✅ Takeaway
Symbolic ordering generalizes classical comparisons. With beta = 0, it collapses to magnitude order; with beta > 0, it exposes stability, preventing unstable numbers from being overvalued.
Navigation
Previous → Primitive objects
Next → Core operations
Disclaimer
Observation only. Reproducible math; domain claims require independent peer review. Defaults: gamma=1, mult_mode=M2, clamp_eps=1e-6, |a|<1.