空間資訊佔據 II:自由度傳輸、邊界飽和與結構逃逸
Spatial Information Occupancy II: Freedom Transport, Boundary Saturation, and Structural Escape
系列: Spatial Information Occupancy and Generator Geometry篇次: Paper 02 / Dynamics版本: v1.0日期: 2026-10-02作者: Neo.KAI 協力研究者: Aletheia / ChatGPT, GPT-5.6 Sol機構: EveMissLab/一言諾科技有限公司狀態: Formal Research Draft / Exact Finite Dynamics
直接前置:
Paper 01《空間資訊佔據 I:超越幾何測度的容器相對自由結構》;
E1 Freedom Transport;
E2 Gap Operator;
主客邊三域代數(SOBTA);
Algebraic–Combinatorial Transcendence Methodology(ACTM)。
摘要
Paper 01 已證明:幾何支撐量並不足以決定容器中的 translation freedom 或 global connectivity effect。本文進一步研究一個動態問題:
當直接佔據量保持不變、幾何結構或邊界開口被逐步改變時,自由度如何在不同 channel 間傳輸、壓縮、釋放,並在何處發生結構逃逸?
本文在有限二維格點容器中建立兩類 exact dynamics。
第一類是固定面積矩形的 aspect-ratio stretch。令 w × h w\times h w × h obstacle 滿足:
w h = K , wh=K, w h = K ,
並以長度 L L L 的 horizontal / vertical probes 測量 placement depletion。在 obstacle 的 probe-influence region 尚未被容器邊界截斷時,本文證明:
I h = h ( w + L − 1 ) , \boxed{
I_h
=
h(w+L-1),
} I h = h ( w + L − 1 ) ,
I v = w ( h + L − 1 ) , \boxed{
I_v
=
w(h+L-1),
} I v = w ( h + L − 1 ) ,
因而:
D v − D h = ( L − 1 ) ( w − h ) N ( N − L + 1 ) . \boxed{
D_v-D_h
=
\frac{(L-1)(w-h)}
{N(N-L+1)}.
} D v − D h = N ( N − L + 1 ) ( L − 1 ) ( w − h ) .
固定 K K K 並沿 w ↑ w\uparrow w ↑ 、 h ↓ h\downarrow h ↓ 的 stretch path,horizontal channel 的 depletion 下降,vertical channel 的 depletion 上升;同時總 placement freedom 下降。因此 deformation 不是單一 scalar freedom 的增減,而是:
channel redistribution + net configuration depletion . \boxed{
\text{channel redistribution}
+
\text{net configuration depletion}.
} channel redistribution + net configuration depletion .
本文再給出一般 boundary clipping 公式。若 obstacle 靠近容器邊界,interior law 必須加入非負 clipping deficits:
β x , β y , \beta_x,
\qquad
\beta_y, β x , β y ,
並有:
I h = h ( w + L − 1 − β x ) , \boxed{
I_h
=
h(w+L-1-\beta_x),
} I h = h ( w + L − 1 − β x ) ,
I v = w ( h + L − 1 − β y ) . \boxed{
I_v
=
w(h+L-1-\beta_y).
} I v = w ( h + L − 1 − β y ) .
因此邊界不是只有發生物理接觸時才重要;它直接修改 admissibility count。
第二類 dynamics 是一格厚 spanning barrier 的 gap operator。對 axis-aligned w p × h p w_p\times h_p w p × h p probe,本文證明:
C a p ( g ; P ) = max ( 0 , g − h p + 1 ) , \boxed{
C_{\mathrm{ap}}(g;P)
=
\max(0,g-h_p+1),
} C ap ( g ; P ) = max ( 0 , g − h p + 1 ) ,
以及 crossing placement 數:
N c r o s s = w p C a p ( g ; P ) . \boxed{
N_{\mathrm{cross}}
=
w_p\,C_{\mathrm{ap}}(g;P).
} N cross = w p C ap ( g ; P ) .
在適當 side-clearance 條件下,left-right configuration-space connectivity 的臨界 gap 精確為:
g c ( P ) = h p . \boxed{
g_c(P)=h_p.
} g c ( P ) = h p .
所以:
geometric opening ≠ operational freedom opening . \boxed{
\text{geometric opening}
\neq
\text{operational freedom opening}.
} geometric opening = operational freedom opening .
當:
g < h p g<h_p g < h p
時,已有 substrate 被釋放,但該 probe 的跨障礙自由仍未形成;本文稱此為 latent freedom 。當 gap operator 把 g = h p − 1 g=h_p-1 g = h p − 1 推至 g = h p g=h_p g = h p 時,configuration space 的 connected-component count 由 2 2 2 跳到 1 1 1 ,形成一個 exact structural escape。
本文因此把 ACTM 的 closure / escape 研究姿勢移植到空間資訊佔據:追蹤哪些 freedom relations 在 operator 下保持、重分配、被邊界修正或離開原 closure class。SOBTA 則被用作關係座標:query/action、constraint object 與 boundary condition 三者共同決定可觀察 freedom,而不是把邊界降格成輪廓附註。
Abstract
Paper 01 established that geometric support alone does not determine translation freedom or global connectivity effects in a finite container. This paper studies the corresponding dynamics: how freedom is redistributed, compressed, released, or structurally transformed when direct occupancy is preserved while shape or boundary openings change.
Two exact finite models are developed.
First, a fixed-area rectangle with w h = K wh=K w h = K is stretched while horizontal and vertical probes of length L L L measure placement depletion. Before the obstacle's probe-influence region is clipped by the container boundary,
I h = h ( w + L − 1 ) , I_h=h(w+L-1), I h = h ( w + L − 1 ) ,
I v = w ( h + L − 1 ) , I_v=w(h+L-1), I v = w ( h + L − 1 ) ,
and therefore
D v − D h = ( L − 1 ) ( w − h ) N ( N − L + 1 ) . \boxed{
D_v-D_h
=
\frac{(L-1)(w-h)}
{N(N-L+1)}.
} D v − D h = N ( N − L + 1 ) ( L − 1 ) ( w − h ) .
Along a fixed-area stretch path with increasing w w w and decreasing h h h , horizontal depletion decreases while vertical depletion increases, whereas total translation freedom decreases. Freedom dynamics are therefore not captured by a single conserved scalar: they exhibit channel redistribution together with net configuration-space depletion.
Second, a one-cell-thick spanning barrier is opened by a contiguous gap of size g g g . For an axis-aligned w p × h p w_p\times h_p w p × h p probe,
C a p ( g ; P ) = max ( 0 , g − h p + 1 ) \boxed{
C_{\mathrm{ap}}(g;P)
=
\max(0,g-h_p+1)
} C ap ( g ; P ) = max ( 0 , g − h p + 1 )
counts admissible transverse aperture positions, and
N c r o s s = w p C a p ( g ; P ) . \boxed{
N_{\mathrm{cross}}
=
w_p C_{\mathrm{ap}}(g;P).
} N cross = w p C ap ( g ; P ) .
Under explicit side-clearance assumptions, left-right connectivity in probe configuration space appears exactly at
g c ( P ) = h p . \boxed{
g_c(P)=h_p.
} g c ( P ) = h p .
Hence a geometrically open region can remain operationally closed for a given action class. The regime 0 < g < h p 0<g<h_p 0 < g < h p is interpreted as latent freedom: released substrate that has not yet formed an admissible crossing.
The paper also derives exact boundary-clipping corrections and formalizes two kinds of structural escape: leaving an interior counting-law class and crossing a configuration-space topological threshold.
1. 從靜態佔據到動態自由度
Paper 01 的基本區分是:
Geometric Measure Occupancy ≠ Freedom Occupancy . \boxed{
\text{Geometric Measure Occupancy}
\neq
\text{Freedom Occupancy}.
} Geometric Measure Occupancy = Freedom Occupancy .
本文進一步考慮一個 operator:
T : ( C , O , q ) → ( C ′ , O ′ , q ′ ) . T:
(\mathfrak C,O,q)
\rightarrow
(\mathfrak C',O',q'). T : ( C , O , q ) → ( C ′ , O ′ , q ′ ) .
研究對象不再只是:
O c c ( O ) , \mathbf{Occ}(O), Occ ( O ) ,
而是:
O c c t → T O c c t + 1 . \boxed{
\mathbf{Occ}_t
\xrightarrow{T}
\mathbf{Occ}_{t+1}.
} Occ t T Occ t + 1 .
本文稱此研究方向為:
Freedom Transport . \boxed{
\text{Freedom Transport}.
} Freedom Transport .
其最低問題包括:
哪些 occupancy 分量被 preserve?
哪些 freedom channel 被 release?
哪些 channel 被 consume?
是否存在 net depletion?
邊界何時改變 counting law?
operator 何時使結構離開原 closure class?
2. 模型 A:固定面積矩形與方向 probe
令容器:
X N = { 0 , … , N − 1 } 2 . X_N
=
\{0,\ldots,N-1\}^2. X N = { 0 , … , N − 1 } 2 .
令 obstacle:
R = [ x 0 , x 0 + w − 1 ] × [ y 0 , y 0 + h − 1 ] R
=
[x_0,x_0+w-1]
\times
[y_0,y_0+h-1] R = [ x 0 , x 0 + w − 1 ] × [ y 0 , y 0 + h − 1 ]
為 axis-aligned w × h w\times h w × h rectangle。
令 horizontal probe:
P h = { 0 , … , L − 1 } × { 0 } , P_h
=
\{0,\ldots,L-1\}\times\{0\}, P h = { 0 , … , L − 1 } × { 0 } ,
vertical probe:
P v = { 0 } × { 0 , … , L − 1 } . P_v
=
\{0\}\times\{0,\ldots,L-1\}. P v = { 0 } × { 0 , … , L − 1 } .
兩者在空容器中的合法 translation anchors 數相同:
Q = N ( N − L + 1 ) . \boxed{
Q
=
N(N-L+1).
} Q = N ( N − L + 1 ) .
定義:
I h = number of horizontal probe anchors invalidated by R , I_h
=
\text{number of horizontal probe anchors invalidated by }R, I h = number of horizontal probe anchors invalidated by R ,
I v = number of vertical probe anchors invalidated by R . I_v
=
\text{number of vertical probe anchors invalidated by }R. I v = number of vertical probe anchors invalidated by R .
相應 depletion:
D h = I h Q , D v = I v Q . \boxed{
D_h=\frac{I_h}{Q},
\qquad
D_v=\frac{I_v}{Q}.
} D h = Q I h , D v = Q I v .
3. Interior placement class
horizontal probe anchor 為:
( a , b ) , (a,b), ( a , b ) ,
其中:
0 ≤ a ≤ N − L , 0 ≤ b ≤ N − 1. 0\le a\le N-L,
\qquad
0\le b\le N-1. 0 ≤ a ≤ N − L , 0 ≤ b ≤ N − 1.
它與 obstacle 相交,當且僅當:
b ∈ [ y 0 , y 0 + h − 1 ] b\in[y_0,y_0+h-1] b ∈ [ y 0 , y 0 + h − 1 ]
且 interval:
[ a , a + L − 1 ] [a,a+L-1] [ a , a + L − 1 ]
與:
[ x 0 , x 0 + w − 1 ] [x_0,x_0+w-1] [ x 0 , x 0 + w − 1 ]
相交。
若 obstacle 在 x x x 方向距左右 probe-anchor clipping boundary 都至少 L − 1 L-1 L − 1 ,則所有理論上會與 obstacle 相交的 horizontal anchors 都仍位於合法 anchor domain 內。
vertical case 類似。
Definition 3.1 — Probe-interior rectangle
若:
x 0 ≥ L − 1 , x_0\ge L-1, x 0 ≥ L − 1 ,
x 0 + w − 1 ≤ N − L , x_0+w-1\le N-L, x 0 + w − 1 ≤ N − L ,
y 0 ≥ L − 1 , y_0\ge L-1, y 0 ≥ L − 1 ,
y 0 + h − 1 ≤ N − L , y_0+h-1\le N-L, y 0 + h − 1 ≤ N − L ,
則稱 R R R 對長度 L L L probe 為 probe-interior 。
這是一個 sufficient condition;後續 boundary theorem 會處理一般情況。
4. Interior Freedom-Transport Law
Theorem 4.1 — Exact Invalid-Placement Counts
若 R R R 為 probe-interior,則:
I h = h ( w + L − 1 ) , \boxed{
I_h
=
h(w+L-1),
} I h = h ( w + L − 1 ) ,
以及:
I v = w ( h + L − 1 ) . \boxed{
I_v
=
w(h+L-1).
} I v = w ( h + L − 1 ) .
Proof
固定 obstacle 中某一 row:
b ∈ [ y 0 , y 0 + h − 1 ] . b\in[y_0,y_0+h-1]. b ∈ [ y 0 , y 0 + h − 1 ] .
horizontal probe 與 obstacle 的 x x x -interval 相交,當且僅當 anchor a a a 落在:
x 0 − L + 1 ≤ a ≤ x 0 + w − 1. x_0-L+1
\le
a
\le
x_0+w-1. x 0 − L + 1 ≤ a ≤ x 0 + w − 1.
probe-interior 條件保證此整段落在:
0 ≤ a ≤ N − L 0\le a\le N-L 0 ≤ a ≤ N − L
內,因此共有:
( x 0 + w − 1 ) − ( x 0 − L + 1 ) + 1 = w + L − 1 (x_0+w-1)-(x_0-L+1)+1
=
w+L-1 ( x 0 + w − 1 ) − ( x 0 − L + 1 ) + 1 = w + L − 1
個 invalid horizontal anchors。
obstacle 有 h h h 個 rows,所以:
I h = h ( w + L − 1 ) . I_h
=
h(w+L-1). I h = h ( w + L − 1 ) .
vertical case 對稱,得到:
I v = w ( h + L − 1 ) . I_v
=
w(h+L-1). I v = w ( h + L − 1 ) .
證畢。
Corollary 4.2 — Directional Anisotropy Law
在 Theorem 4.1 條件下:
I v − I h = ( L − 1 ) ( w − h ) . \boxed{
I_v-I_h
=
(L-1)(w-h).
} I v − I h = ( L − 1 ) ( w − h ) .
因此:
D v − D h = ( L − 1 ) ( w − h ) N ( N − L + 1 ) . \boxed{
D_v-D_h
=
\frac{
(L-1)(w-h)
}{
N(N-L+1)
}.
} D v − D h = N ( N − L + 1 ) ( L − 1 ) ( w − h ) .
Proof
直接相減:
I v − I h = w ( h + L − 1 ) − h ( w + L − 1 ) = w h + w ( L − 1 ) − h w − h ( L − 1 ) = ( L − 1 ) ( w − h ) . \begin{aligned}
I_v-I_h
&=
w(h+L-1)
-
h(w+L-1)\\
&=
wh+w(L-1)-hw-h(L-1)\\
&=
(L-1)(w-h).
\end{aligned} I v − I h = w ( h + L − 1 ) − h ( w + L − 1 ) = w h + w ( L − 1 ) − h w − h ( L − 1 ) = ( L − 1 ) ( w − h ) .
再除以共同 denominator Q Q Q 。證畢。
5. 固定 direct support 下的 freedom redistribution
現在固定:
w h = K . wh=K. w h = K .
在 interior class 中:
I h = K + h ( L − 1 ) , I_h
=
K+h(L-1), I h = K + h ( L − 1 ) ,
I v = K + w ( L − 1 ) . I_v
=
K+w(L-1). I v = K + w ( L − 1 ) .
所以:
D h = K + h ( L − 1 ) Q , \boxed{
D_h
=
\frac{
K+h(L-1)
}{
Q
},
} D h = Q K + h ( L − 1 ) ,
D v = K + w ( L − 1 ) Q . \boxed{
D_v
=
\frac{
K+w(L-1)
}{
Q
}.
} D v = Q K + w ( L − 1 ) .
Theorem 5.1 — Fixed-Area Directional Redistribution
令兩個 probe-interior rectangles:
R 1 = ( w 1 , h 1 ) , R 2 = ( w 2 , h 2 ) R_1=(w_1,h_1),
\qquad
R_2=(w_2,h_2) R 1 = ( w 1 , h 1 ) , R 2 = ( w 2 , h 2 )
滿足:
w 1 h 1 = w 2 h 2 = K , w_1h_1=w_2h_2=K, w 1 h 1 = w 2 h 2 = K ,
以及:
w 2 > w 1 . w_2>w_1. w 2 > w 1 .
則:
h 2 < h 1 , h_2<h_1, h 2 < h 1 ,
且:
D h ( R 2 ) < D h ( R 1 ) , \boxed{
D_h(R_2)<D_h(R_1),
} D h ( R 2 ) < D h ( R 1 ) ,
D v ( R 2 ) > D v ( R 1 ) . \boxed{
D_v(R_2)>D_v(R_1).
} D v ( R 2 ) > D v ( R 1 ) .
Proof
固定:
w h = K , wh=K, w h = K ,
故 w w w 增加時 h = K / w h=K/w h = K / w 減少。
而:
D h = K + ( L − 1 ) h Q D_h
=
\frac{
K+(L-1)h
}{Q} D h = Q K + ( L − 1 ) h
對 h h h 嚴格遞增,所以 h h h 下降使 D h D_h D h 下降。
同理:
D v = K + ( L − 1 ) w Q D_v
=
\frac{
K+(L-1)w
}{Q} D v = Q K + ( L − 1 ) w
對 w w w 嚴格遞增,因此 w w w 增加使 D v D_v D v 上升。證畢。
Corollary 5.2 — Mean Probe Depletion
定義 mean directional depletion:
M = D h + D v 2 . \boxed{
M
=
\frac{D_h+D_v}{2}.
} M = 2 D h + D v .
則:
M = K + L − 1 2 ( w + h ) Q . \boxed{
M
=
\frac{
K+\frac{L-1}{2}(w+h)
}{
Q
}.
} M = Q K + 2 L − 1 ( w + h ) .
固定 K K K 時,由 AM-GM:
w + h ≥ 2 K . w+h
\ge
2\sqrt K. w + h ≥ 2 K .
因此正方形:
w = h = K w=h=\sqrt K w = h = K
最小化:
M . M. M .
也就是說,在這個 probe model 中,離開 square symmetry 不只是把 freedom 從一個方向搬到另一個方向;平均 probe depletion 也會上升。
6. Translation freedom 與 net configuration depletion
Paper 01 已證明 rectangle 的合法 translation anchors:
P N ( w , h ) = ( N − w + 1 ) ( N − h + 1 ) . P_N(w,h)
=
(N-w+1)(N-h+1). P N ( w , h ) = ( N − w + 1 ) ( N − h + 1 ) .
固定:
w h = K wh=K w h = K
後:
P N ( w , h ) = ( N + 1 ) 2 + K − ( N + 1 ) ( w + h ) . \boxed{
P_N(w,h)
=
(N+1)^2+K-(N+1)(w+h).
} P N ( w , h ) = ( N + 1 ) 2 + K − ( N + 1 ) ( w + h ) .
若以:
N 2 N^2 N 2
作為固定 ambient anchor normalization,定義:
F t r a n s = P N ( w , h ) N 2 . \boxed{
F_{\mathrm{trans}}
=
\frac{
P_N(w,h)
}{
N^2
}.
} F trans = N 2 P N ( w , h ) .
因此 square 最大化:
F t r a n s . F_{\mathrm{trans}}. F trans .
結合 Theorem 5.1,可得到:
Corollary 6.1 — Redistribution with Net Depletion
沿 fixed-area path 從 square 往:
w > h w>h w > h
方向拉長時:
D h ↓ , \boxed{
D_h\downarrow,
} D h ↓ ,
D v ↑ , \boxed{
D_v\uparrow,
} D v ↑ ,
但:
F t r a n s ↓ . \boxed{
F_{\mathrm{trans}}\downarrow.
} F trans ↓ .
所以:
Freedom Transport ≠ simple scalar conservation . \boxed{
\text{Freedom Transport}
\neq
\text{simple scalar conservation}.
} Freedom Transport = simple scalar conservation .
在本模型中,更準確的是:
directional redistribution + net placement depletion . \boxed{
\text{directional redistribution}
+
\text{net placement depletion}.
} directional redistribution + net placement depletion .
7. Boundary clipping 的一般公式
interior law 不是所有 obstacle positions 都成立。
令:
c x c_x c x
為 horizontal probe anchor 在 x x x 方向上因 obstacle 失效的合法 anchor 數。
由 interval intersection:
c x = max ( 0 , min ( N − L , x 0 + w − 1 ) − max ( 0 , x 0 − L + 1 ) + 1 ) . \boxed{
c_x
=
\max
\left(
0,
\min(N-L,x_0+w-1)
-
\max(0,x_0-L+1)
+
1
\right).
} c x = max ( 0 , min ( N − L , x 0 + w − 1 ) − max ( 0 , x 0 − L + 1 ) + 1 ) .
同理:
c y = max ( 0 , min ( N − L , y 0 + h − 1 ) − max ( 0 , y 0 − L + 1 ) + 1 ) . \boxed{
c_y
=
\max
\left(
0,
\min(N-L,y_0+h-1)
-
\max(0,y_0-L+1)
+
1
\right).
} c y = max ( 0 , min ( N − L , y 0 + h − 1 ) − max ( 0 , y 0 − L + 1 ) + 1 ) .
Theorem 7.1 — Exact Boundary-Clipped Probe Counts
對任意合法 obstacle rectangle:
I h = h c x , \boxed{
I_h
=
h\,c_x,
} I h = h c x ,
I v = w c y . \boxed{
I_v
=
w\,c_y.
} I v = w c y .
Proof
horizontal probe 只有在其 row 落於 obstacle 的 h h h 個 occupied rows 時才可能與 obstacle 相交。
對每一 occupied row,可造成相交的合法 x x x anchors 恰好是理論 overlap interval:
[ x 0 − L + 1 , x 0 + w − 1 ] [x_0-L+1,x_0+w-1] [ x 0 − L + 1 , x 0 + w − 1 ]
與合法 anchor interval:
[ 0 , N − L ] [0,N-L] [ 0 , N − L ]
的交集。
其 integer cardinality 就是 c x c_x c x 。
因此:
I h = h c x . I_h=h\,c_x. I h = h c x .
vertical case 對稱。證畢。
8. Boundary Deficit
定義 interior-theory anchor span:
w + L − 1 w+L-1 w + L − 1
與:
h + L − 1. h+L-1. h + L − 1.
再定義 boundary clipping deficits:
β x = ( w + L − 1 ) − c x , \boxed{
\beta_x
=
(w+L-1)-c_x,
} β x = ( w + L − 1 ) − c x ,
β y = ( h + L − 1 ) − c y . \boxed{
\beta_y
=
(h+L-1)-c_y.
} β y = ( h + L − 1 ) − c y .
由 c x ≤ w + L − 1 c_x\le w+L-1 c x ≤ w + L − 1 與 c y ≤ h + L − 1 c_y\le h+L-1 c y ≤ h + L − 1 :
β x , β y ≥ 0. \beta_x,\beta_y\ge0. β x , β y ≥ 0.
Theorem 7.1 可重寫成:
I h = h ( w + L − 1 − β x ) , \boxed{
I_h
=
h(w+L-1-\beta_x),
} I h = h ( w + L − 1 − β x ) ,
I v = w ( h + L − 1 − β y ) . \boxed{
I_v
=
w(h+L-1-\beta_y).
} I v = w ( h + L − 1 − β y ) .
因此:
Corollary 8.1 — Boundary-Corrected Anisotropy
I v − I h = ( L − 1 ) ( w − h ) + h β x − w β y . \boxed{
I_v-I_h
=
(L-1)(w-h)
+
h\beta_x
-
w\beta_y.
} I v − I h = ( L − 1 ) ( w − h ) + h β x − w β y .
這個式子把 anisotropy 分成:
intrinsic shape term:
( L − 1 ) ( w − h ) , (L-1)(w-h), ( L − 1 ) ( w − h ) ,
boundary correction:
h β x − w β y . h\beta_x-w\beta_y. h β x − w β y .
因此:
observed directional depletion = shape contribution + boundary contribution . \boxed{
\text{observed directional depletion}
=
\text{shape contribution}
+
\text{boundary contribution}.
} observed directional depletion = shape contribution + boundary contribution .
9. Boundary Saturation Example
令:
N = 64 , N=64, N = 64 ,
L = 8 , L=8, L = 8 ,
w = 64 , w=64, w = 64 ,
h = 9. h=9. h = 9.
obstacle 在 x x x 方向跨完整容器,因此:
x 0 = 0. x_0=0. x 0 = 0.
此時:
c x = N − L + 1 = 57. c_x
=
N-L+1
=
57. c x = N − L + 1 = 57.
而 interior theory 會給:
w + L − 1 = 71. w+L-1
=
71. w + L − 1 = 71.
因此:
β x = 14. \beta_x
=
14. β x = 14.
若 y y y 方向仍 probe-interior,則:
β y = 0. \beta_y=0. β y = 0.
所以:
I h = 9 ( 71 − 14 ) = 513 , I_h
=
9(71-14)
=
513, I h = 9 ( 71 − 14 ) = 513 ,
I v = 64 ( 9 + 8 − 1 ) = 1024. I_v
=
64(9+8-1)
=
1024. I v = 64 ( 9 + 8 − 1 ) = 1024.
相對 interior continuation,horizontal residual:
R h = − 9 ⋅ 14 = − 126. R_h
=
-9\cdot14
=
-126. R h = − 9 ⋅ 14 = − 126.
這不是公式「近似失準」,而是合法 anchor domain 被 container boundary 截斷。
因此 boundary saturation 應被理解為:
change of admissibility counting law . \boxed{
\text{change of admissibility counting law}.
} change of admissibility counting law .
10. 模型 B:Spanning Barrier 與 Gap Operator
令:
B g B_g B g
是一格厚 vertical barrier,位於:
x = c , x=c, x = c ,
並具有一段連續 vertical gap:
G g = { y 0 , … , y 0 + g − 1 } . G_g
=
\{y_0,\ldots,y_0+g-1\}. G g = { y 0 , … , y 0 + g − 1 } .
除了 gap 外:
( c , y ) (c,y) ( c , y )
皆被 barrier 佔據。
令 probe:
P = w p × h p P
=
w_p\times h_p P = w p × h p
為 axis-aligned rectangle。
probe translation anchor:
( a , b ) (a,b) ( a , b )
代表其支撐:
[ a , a + w p − 1 ] × [ b , b + h p − 1 ] . [a,a+w_p-1]
\times
[b,b+h_p-1]. [ a , a + w p − 1 ] × [ b , b + h p − 1 ] .
11. Transverse Aperture Capacity
probe placement 若要在跨 barrier 的同時不碰到 barrier,則:
[ b , b + h p − 1 ] [b,b+h_p-1] [ b , b + h p − 1 ]
必須完全包含於 gap:
G g . G_g. G g .
Theorem 11.1 — Exact Aperture Capacity
vertical gap 中可容納 probe transverse span 的合法 b b b anchors 數為:
C a p ( g ; P ) = max ( 0 , g − h p + 1 ) . \boxed{
C_{\mathrm{ap}}(g;P)
=
\max(0,g-h_p+1).
} C ap ( g ; P ) = max ( 0 , g − h p + 1 ) .
Proof
若:
g < h p , g<h_p, g < h p ,
沒有任何長度 h p h_p h p 的 integer interval 能完全放入長度 g g g 的 gap,因此:
C a p = 0. C_{\mathrm{ap}}=0. C ap = 0.
若:
g ≥ h p , g\ge h_p, g ≥ h p ,
probe vertical anchor b b b 可從 gap 起點:
y 0 y_0 y 0
到:
y 0 + g − h p y_0+g-h_p y 0 + g − h p
共:
g − h p + 1 g-h_p+1 g − h p + 1
個位置。證畢。
12. Crossing Placement Count
若 probe 要與 barrier column:
x = c x=c x = c
相交,horizontal anchor a a a 必須滿足:
a ≤ c ≤ a + w p − 1. a\le c\le a+w_p-1. a ≤ c ≤ a + w p − 1.
若 barrier 距左右 container edge 有足夠 clearance,使整個 anchor range 都合法,則:
a a a
共有:
w p w_p w p
個值。
Corollary 12.1 — Exact Cross-Barrier Placement Count
在 side-clearance 條件下:
N c r o s s ( g ; P ) = w p max ( 0 , g − h p + 1 ) . \boxed{
N_{\mathrm{cross}}(g;P)
=
w_p
\max(0,g-h_p+1).
} N cross ( g ; P ) = w p max ( 0 , g − h p + 1 ) .
所以:
N c r o s s > 0 ⟺ g ≥ h p . \boxed{
N_{\mathrm{cross}}>0
\iff
g\ge h_p.
} N cross > 0 ⟺ g ≥ h p .
13. Configuration-Space Connectivity Threshold
令:
Q P ( B g ) \mathcal Q_P(B_g) Q P ( B g )
為所有合法 probe translation anchors 構成的 graph,兩個 anchors 若相差一個 unit horizontal / vertical translation 則相鄰。
假設:
probe 在 barrier 左右兩側都存在合法 placements;
gap 不與 container 上下 boundary 相衝突;
barrier 為單一 spanning vertical strip;
probe 不允許旋轉,只做 axis-aligned translation。
Theorem 13.1 — Critical Gap Theorem
在上述條件下:
# π 0 ( Q P ( B g ) ) = { 2 , g < h p , 1 , g ≥ h p . \boxed{
\#\pi_0(\mathcal Q_P(B_g))
=
\begin{cases}
2,
&
g<h_p,
\\
1,
&
g\ge h_p.
\end{cases}
} # π 0 ( Q P ( B g )) = { 2 , 1 , g < h p , g ≥ h p .
因此 critical gap:
g c ( P ) = h p . \boxed{
g_c(P)=h_p.
} g c ( P ) = h p .
Proof
Case 1: g < h p g<h_p g < h p
由 Theorem 11.1:
C a p = 0. C_{\mathrm{ap}}=0. C ap = 0.
因此不存在任何合法 placement 同時跨越 barrier column。
所有合法 anchors 只能位於 barrier 左側或右側。
因 barrier 從上到下 spanning container,且 probe 無法穿越 gap,所以左側與右側 anchor sets 間不存在 configuration-space path。
每一側本身為 rectangular / orthogonally connected anchor region,因此 configuration graph 恰有兩個 connected components。
Case 2: g ≥ h p g\ge h_p g ≥ h p
由 Theorem 11.1:
C a p ≥ 1. C_{\mathrm{ap}}\ge1. C ap ≥ 1.
所以存在某個 vertical anchor b ∗ b^\ast b ∗ ,使 probe entire vertical footprint 落在 gap 內。
由 side-clearance 條件,存在一串 horizontal anchors:
a L , a L + 1 , … , a R a_L,
a_L+1,
\ldots,
a_R a L , a L + 1 , … , a R
使 probe 從 barrier 左側逐步 translation,經過跨 barrier placements,再到 barrier 右側,且整段都保持合法。
左右側任意合法 placement 可先在各自 free rectangular anchor region 內移動到此 corridor,再穿越至另一側。
因此整個:
Q P ( B g ) \mathcal Q_P(B_g) Q P ( B g )
連通。
證畢。
14. Boundary Permeability
Theorem 13.1 允許定義最小 binary permeability:
κ P ( g ) = 1 [ g ≥ h p ] . \boxed{
\kappa_P(g)
=
\mathbf 1[g\ge h_p].
} κ P ( g ) = 1 [ g ≥ h p ] .
以及更細的 aperture capacity:
C a p ( g ; P ) . C_{\mathrm{ap}}(g;P). C ap ( g ; P ) .
因此同一 boundary opening g g g ,對不同 probe:
P 1 , P 2 P_1,
\qquad
P_2 P 1 , P 2
可以有:
κ P 1 ( g ) ≠ κ P 2 ( g ) . \kappa_{P_1}(g)
\neq
\kappa_{P_2}(g). κ P 1 ( g ) = κ P 2 ( g ) .
這不是 observer arbitrary choice,而是不同 action class 的 admissibility structure 不同。
15. Released Substrate、Latent Freedom 與 Realized Freedom
Theorem 11.1 顯示:
0 < g < h p 0<g<h_p 0 < g < h p
時:
g > 0 , g>0, g > 0 ,
但:
C a p = 0. C_{\mathrm{ap}}=0. C ap = 0.
因此需要區分三層:
Released Substrate → Latent Freedom → Realized Operational Freedom . \boxed{
\text{Released Substrate}
\rightarrow
\text{Latent Freedom}
\rightarrow
\text{Realized Operational Freedom}.
} Released Substrate → Latent Freedom → Realized Operational Freedom .
Definition 15.1 — Gap Deficit
對 probe P P P 定義:
δ P ( g ) = max ( 0 , h p − g ) . \boxed{
\delta_P(g)
=
\max(0,h_p-g).
} δ P ( g ) = max ( 0 , h p − g ) .
若:
δ P ( g ) = 0 , \delta_P(g)=0, δ P ( g ) = 0 ,
表示 crossing freedom 已存在。
若:
δ P ( g ) = k > 0 , \delta_P(g)=k>0, δ P ( g ) = k > 0 ,
表示至少還需要額外釋放 k k k 個連續 transverse constituents 才能達到 crossing threshold。
這是 Paper 03 更一般 latent-freedom potential:
d q ( A ) d_q(A) d q ( A )
的最簡一維前身。
16. Local Release 與 Global Topological Jump
Gap operator:
G : B g ↦ B g + 1 G:
B_g
\mapsto
B_{g+1} G : B g ↦ B g + 1
每次只釋放一個 barrier site。
因此:
Δ η s i t e = − 1 N 2 . \boxed{
\Delta\eta_{\mathrm{site}}
=
-\frac{1}{N^2}.
} Δ η site = − N 2 1 .
但當:
g = h p − 1 g=h_p-1 g = h p − 1
推進到:
g = h p , g=h_p, g = h p ,
Theorem 13.1 給出:
# π 0 : 2 → 1. \boxed{
\#\pi_0:
2
\rightarrow
1.
} # π 0 : 2 → 1.
Corollary 16.1 — Vanishing Local Measure, Persistent Topological Transition
取一族 container sizes:
N → ∞ N\rightarrow\infty N → ∞
與 probes:
h p ( N ) h_p(N) h p ( N )
使 side-clearance 條件保持,並讓 transition 每次仍由:
g = h p ( N ) − 1 → h p ( N ) g=h_p(N)-1
\rightarrow
h_p(N) g = h p ( N ) − 1 → h p ( N )
觸發。
則單步 direct occupancy change:
∣ Δ η s i t e ∣ = 1 N 2 → 0 , \boxed{
\left|
\Delta\eta_{\mathrm{site}}
\right|
=
\frac{1}{N^2}
\rightarrow0,
} ∣ Δ η site ∣ = N 2 1 → 0 ,
但 configuration-space topological change:
Δ # π 0 = − 1 \boxed{
\Delta\#\pi_0=-1
} Δ# π 0 = − 1
保持不消失。
因此:
vanishing local geometric change ⇏ vanishing global freedom-structure change . \boxed{
\text{vanishing local geometric change}
\not\Rightarrow
\text{vanishing global freedom-structure change}.
} vanishing local geometric change ⇒ vanishing global freedom-structure change .
17. 兩種 Structural Escape
本文區分兩種不同 escape。
17.1 Counting-Law Escape
定義 probe-interior class:
K i n t \mathcal K_{\mathrm{int}} K int
為:
β x = β y = 0 \beta_x=\beta_y=0 β x = β y = 0
的 obstacle states。
若 deformation operator:
T T T
使:
β x > 0 \beta_x>0 β x > 0
或:
β y > 0 , \beta_y>0, β y > 0 ,
則:
T ( O ) ∉ K i n t . \boxed{
T(O)
\notin
\mathcal K_{\mathrm{int}}.
} T ( O ) ∈ / K int .
原 interior counting law 必須改用 boundary-corrected law。
17.2 Topological Escape
對固定 probe:
K s p l i t ( P ) = { B g : g < h p } . \mathcal K_{\mathrm{split}}^{(P)}
=
\{
B_g:
g<h_p
\}. K split ( P ) = { B g : g < h p } .
gap operator:
G : B g ↦ B g + 1 G:
B_g
\mapsto
B_{g+1} G : B g ↦ B g + 1
在:
g = h p − 1 g=h_p-1 g = h p − 1
處造成:
G ( B h p − 1 ) ∉ K s p l i t ( P ) , \boxed{
G(B_{h_p-1})
\notin
\mathcal K_{\mathrm{split}}^{(P)},
} G ( B h p − 1 ) ∈ / K split ( P ) ,
並伴隨:
# π 0 : 2 → 1. \#\pi_0:
2\rightarrow1. # π 0 : 2 → 1.
第一種 escape 改變 counting regime。
第二種 escape 改變 configuration-space topology。
兩者不可混同。
18. ACTM-style Freedom Audit
對 fixed-area stretch:
T s t r e t c h : ( w , h ) ↦ ( w ′ , h ′ ) T_{\mathrm{stretch}}
:
(w,h)
\mapsto
(w',h') T stretch : ( w , h ) ↦ ( w ′ , h ′ )
且:
w h = w ′ h ′ = K , wh=w'h'=K, w h = w ′ h ′ = K ,
可做:
Preserve
K , η s i t e . K,
\qquad
\eta_{\mathrm{site}}. K , η site .
Release
在 w w w 增加的 path:
D h ↓ . D_h\downarrow. D h ↓ .
Consume
D v ↑ . D_v\uparrow. D v ↑ .
Compress
F t r a n s ↓ . F_{\mathrm{trans}}\downarrow. F trans ↓ .
Escape
若 boundary clipping deficit 由:
0 0 0
變成正值,離開 interior closure class。
對 gap operator:
Release substrate
∣ B g ∣ ↓ . |B_g|
\downarrow. ∣ B g ∣ ↓ .
Accumulate latent freedom
δ P ( g ) ↓ . \delta_P(g)\downarrow. δ P ( g ) ↓ .
Realize freedom
在:
g = h p g=h_p g = h p
時:
κ P : 0 → 1. \kappa_P:
0\rightarrow1. κ P : 0 → 1.
Topological escape
# π 0 : 2 → 1. \#\pi_0:
2\rightarrow1. # π 0 : 2 → 1.
這就是本文所謂 ACTM-style freedom transport:不是把 ACTM 的超越數工具直接搬到幾何,而是移植其「追蹤 relation 在 operator 下的 closure / escape」方法。
19. SOBTA:Boundary as Admissibility Coupling
本文使用 SOBTA:
Γ q = ⟨ S q , O , ∂ C ⟩ \Gamma_q
=
\langle
S_q,O,\partial\mathfrak C
\rangle Γ q = ⟨ S q , O , ∂ C ⟩
作為 relational coordinate。
其中:
S q S_q S q
指定:
probe;
action class;
configuration adjacency;
resolution;
query semantics。
O O O
指定 obstacle / constraint。
∂ C \partial\mathfrak C ∂ C
決定:
anchor clipping;
crossing 是否可行;
哪些 deformation 留在合法 domain;
哪些局部 opening 可以轉成全域 transport。
因此本文的 exact formulas 顯示:
freedom observable = f ( S q , O , ∂ C ) . \boxed{
\text{freedom observable}
=
f(
S_q,
O,
\partial\mathfrak C
).
} freedom observable = f ( S q , O , ∂ C ) .
但這不推出:
all geometry is ontologically SOBTA . \boxed{
\text{all geometry is ontologically SOBTA}.
} all geometry is ontologically SOBTA .
SOBTA 在此是 coordinate framework,而不是 reduction claim。
20. 「邊界」的兩個不同層次
E1/E2 迫使我們把 boundary 至少分成:
20.1 Container Boundary
∂ C \partial\mathfrak C ∂ C
決定 legal anchor domain。
其作用由:
β x , β y \beta_x,
\beta_y β x , β y
顯示。
20.2 Internal Barrier Boundary
∂ O \partial O ∂ O
決定 configuration space 是否分裂、gap 是否具有 permeability。
因此:
boundary ≠ single geometric contour notion . \boxed{
\text{boundary}
\neq
\text{single geometric contour notion}.
} boundary = single geometric contour notion .
邊界更一般地是:
structure that changes admissibility relations . \boxed{
\text{structure that changes admissibility relations}.
} structure that changes admissibility relations .
這個定義在本文中是 operational,而不是終極本體定義。
21. Computational Audit
本文主要定理都有解析證明。
附帶:
paper02_exact_checks.py
只作 proof audit,驗證:
interior probe-count formulas;
fixed-area directional monotonicity;
boundary clipping general formula;
N = 64 , L = 8 , w = 64 , h = 9 N=64,L=8,w=64,h=9 N = 64 , L = 8 , w = 64 , h = 9 的 − 126 -126 − 126 horizontal residual;
gap capacity;
cross-barrier placement count;
critical gap;
configuration component transition。
程式驗證不取代解析證明。
22. 可證偽性與限制
F1 — Boundary Formula Failure
若 Theorem 7.1 在明確 rectangle / probe model 中失敗,本文 boundary interpretation 失效。
F2 — Threshold Failure
若存在符合本文 side-clearance 假設的 probe / barrier,卻有:
g c ≠ h p , g_c\neq h_p, g c = h p ,
則 Theorem 13.1 失效。
F3 — Query Inflation
若任意改 query 都可事後合理化結果,框架失去辨識力。
因此 query 必須在計算前固定 probe、action、adjacency 與 admissibility。
F4 — Overgeneralization
本文不宣稱:
g c = h p g_c=h_p g c = h p
適用於所有 curved barrier、rotation-enabled probe 或 continuous deformation。
它只適用本文明確的 axis-aligned translation model。
F5 — Physical Ontology Overreach
本文不從 finite lattice topology jump 推出宇宙物理相變。
它只證明:
在明確的有限 configuration-space model 中,局部 direct-support change 可以對 global freedom topology 造成非比例影響。
23. 與下一篇的接口
本文得到:
0 < g < h p 0<g<h_p 0 < g < h p
時 substrate 已釋放,但 operational crossing 尚未形成。
這立刻留下問題:
那些尚未閉合成 freedom 的 constituents 到底承載了什麼?
Paper 03 將把 gap 從單一連續區段推廣到任意 released subset:
A ⊆ W . A\subseteq W. A ⊆ W .
並建立:
C h ( A ) = ∑ J ∈ W h 1 [ J ⊆ A ] , C_h(A)
=
\sum_{J\in\mathcal W_h}
\mathbf 1[J\subseteq A], C h ( A ) = J ∈ W h ∑ 1 [ J ⊆ A ] ,
以及:
d q ( A ) = min G ∈ G q ∣ G ∖ A ∣ . d_q(A)
=
\min_{G\in\mathcal G_q}
|G\setminus A|. d q ( A ) = G ∈ G q min ∣ G ∖ A ∣.
由此研究:
cooperative constituent information;
Möbius / Harsanyi interaction;
Shapley field;
latent freedom potential;
minimal generator hypergraph。
24. 結論
本文把 Paper 01 的靜態區分:
measure ≠ freedom \text{measure}
\neq
\text{freedom} measure = freedom
推進為動態結構。
固定 direct support:
w h = K wh=K w h = K
並不代表 freedom 不變。
相反地:
D h ↓ , D v ↑ , F t r a n s ↓ \boxed{
D_h\downarrow,
\qquad
D_v\uparrow,
\qquad
F_{\mathrm{trans}}\downarrow
} D h ↓ , D v ↑ , F trans ↓
可以同時成立。
這表示 deformation 可以:
redistribute freedom + deplete total configuration capacity . \boxed{
\text{redistribute freedom}
+
\text{deplete total configuration capacity}.
} redistribute freedom + deplete total configuration capacity .
當 obstacle 接近 container boundary:
β x , β y \boxed{
\beta_x,\beta_y
} β x , β y
開始修改 counting law。
當 barrier gap 達到 probe transverse thickness:
g = h p , g=h_p, g = h p ,
則:
latent freedom → realized operational freedom , \boxed{
\text{latent freedom}
\rightarrow
\text{realized operational freedom},
} latent freedom → realized operational freedom ,
且 configuration space 發生:
# π 0 : 2 → 1. \boxed{
\#\pi_0:
2\rightarrow1.
} # π 0 : 2 → 1.
因此邊界在本文中的成熟 operational 定義是:
a structure that modifies admissibility laws and freedom connectivity. \boxed{
\textbf{a structure that modifies admissibility laws and freedom connectivity.}
} a structure that modifies admissibility laws and freedom connectivity.
而一句話總結本文:
Freedom is not merely lost or preserved; it can be transported across channels and can change regime at boundaries. \boxed{
\textbf{Freedom is not merely lost or preserved; it can be transported across channels and can change regime at boundaries.}
} Freedom is not merely lost or preserved; it can be transported across channels and can change regime at boundaries.
中文:
自由度不只是被保留或被消耗;它可以在不同 channel 間傳輸,並在邊界上改變其合法作用規則。
Appendix A. 核心公式
Interior counts:
I h = h ( w + L − 1 ) , I_h=h(w+L-1), I h = h ( w + L − 1 ) ,
I v = w ( h + L − 1 ) . I_v=w(h+L-1). I v = w ( h + L − 1 ) .
Directional anisotropy:
D v − D h = ( L − 1 ) ( w − h ) N ( N − L + 1 ) . D_v-D_h
=
\frac{(L-1)(w-h)}
{N(N-L+1)}. D v − D h = N ( N − L + 1 ) ( L − 1 ) ( w − h ) .
Boundary-clipped counts:
I h = h c x , I_h=h\,c_x, I h = h c x ,
I v = w c y . I_v=w\,c_y. I v = w c y .
Clipping deficits:
β x = ( w + L − 1 ) − c x , \beta_x=(w+L-1)-c_x, β x = ( w + L − 1 ) − c x ,
β y = ( h + L − 1 ) − c y . \beta_y=(h+L-1)-c_y. β y = ( h + L − 1 ) − c y .
Boundary-corrected anisotropy:
I v − I h = ( L − 1 ) ( w − h ) + h β x − w β y . I_v-I_h
=
(L-1)(w-h)
+
h\beta_x
-
w\beta_y. I v − I h = ( L − 1 ) ( w − h ) + h β x − w β y .
Gap capacity:
C a p ( g ; P ) = max ( 0 , g − h p + 1 ) . C_{\mathrm{ap}}(g;P)
=
\max(0,g-h_p+1). C ap ( g ; P ) = max ( 0 , g − h p + 1 ) .
Cross-barrier placements:
N c r o s s = w p C a p . N_{\mathrm{cross}}
=
w_p C_{\mathrm{ap}}. N cross = w p C ap .
Critical gap:
g c ( P ) = h p . g_c(P)=h_p. g c ( P ) = h p .
Gap deficit:
δ P ( g ) = max ( 0 , h p − g ) . \delta_P(g)
=
\max(0,h_p-g). δ P ( g ) = max ( 0 , h p − g ) .
Appendix B. Formal Status
解析定理:
Theorem 4.1;
Theorem 5.1;
Theorem 7.1;
Theorem 11.1;
Theorem 13.1;
Corollaries 4.2、5.2、6.1、8.1、12.1、16.1。
Computational audit only:
finite parameter sweeps;
example counts;
BFS connectivity cross-checks。
尚未宣稱:
arbitrary rotation theorem;
arbitrary curved barrier theorem;
continuous physical-space theorem;
general topological phase-transition universality;
universal freedom conservation law。