feat: deploy Kalpana RIF O(1) Studio with Needle-in-a-Haystack benchmarks, layer architecture, and Swagger API

README.md

@@ -21,9 +21,11 @@ license: apache-2.0
 
 ## 🏛️ Architecture Overview
 
-Kalpana RIF replaces standard linear $O(N)$ Key-Value tensor caching in Transformer LLMs with continuous wave interference states governed by:
+Kalpana RIF replaces standard linear $O(N)$ Key-Value tensor caching in Transformer LLMs across all 32 hidden layers with proprietary bounded-memory continuous state matrices:
 
-$$\Psi(t) = \Psi(t-1) + \kappa \sum_{b=1}^{B} \Big[ \cos(\omega_b t + \phi_b) + i \sin(\omega_b t + \phi_b) \Big] \mathbf{v}_t$$
+- **Drop-in Attention Layer Replacement:** Intercepts Key ($K$) and Value ($V$) states across all attention heads.
+- **Strict $O(1)$ Invariant Footprint:** Preserves constant memory whether processing 2K or 3,000,000 tokens.
+- **Zero Prompt Recomputation:** Instant multi-turn recall with bounded compute cost.
 
 ---
 

index.html

@@ -283,7 +283,7 @@
               <div class="diagram-arrow">▼</div>
 
               <div class="diagram-block block-transformer">
-                <div class="block-title">2. Transformer Hidden Layers ($L = 0 \dots 31$)</div>
+                <div class="block-title">2. Transformer Hidden Layer Stack ($L = 0 \dots 31$)</div>
                 <div class="block-desc">Every layer contains Multi-Head Self Attention (32 parallel heads). Each head projects Query ($Q_t$), Key ($K_t$), and Value ($V_t$).</div>
 
                 <!-- Inner Interception Layer -->
@@ -292,14 +292,14 @@
                   
                   <div class="interception-grid">
                     <div class="sub-block">
-                      <strong>Standard KV (Eliminated):</strong>
+                      <strong>Standard Transformers KV:</strong>
                       <code>torch.cat([Buffer_{t-1}, K_t], dim=-2)</code>
-                      <span class="val-rose">❌ O(N) Unbounded VRAM</span>
+                      <span class="val-rose">❌ Linear O(N) Unbounded VRAM</span>
                     </div>
                     <div class="sub-block">
-                      <strong>Kalpana RIF (Active):</strong>
-                      <code>\Psi(t) += \kappa \cos(\omega_b t + \phi_b) \mathbf{v}_t</code>
-                      <span class="val-emerald">✅ Strict O(1) 6.00 MB</span>
+                      <strong>Kalpana RIF Attention Substrate:</strong>
+                      <code>KalpanaCacheLayer(Past Key-Values)</code>
+                      <span class="val-emerald">✅ Strict O(1) Constant Memory</span>
                     </div>
                   </div>
                 </div>
@@ -308,41 +308,58 @@
               <div class="diagram-arrow">▼</div>
 
               <div class="diagram-block block-sweep">
-                <div class="block-title">3. Holographic Phase Reconstruction & Attention</div>
+                <div class="block-title">3. Holographic Phase Reconstruction & Attention Output</div>
                 <div class="block-desc">
-                  When Query attends to context, continuous harmonic frequencies reconstruct $\hat{K}$ and $\hat{V}$ via phase resonance:
-                  $$\text{Attention}(Q, K, V) = \text{softmax}\left(\frac{Q \hat{K}^T}{\sqrt{d}}\right) \hat{V}$$
+                  When Query attends to context, continuous harmonic wave channels reconstruct Key and Value states deterministically for attention synthesis.
                 </div>
               </div>
 
               <div class="diagram-arrow">▼</div>
 
               <div class="diagram-block block-output">
-                <div class="block-title">4. Autoregressive Output Token</div>
-                <div class="block-desc">Next token prediction with zero context length recomputation tax.</div>
+                <div class="block-title">4. Autoregressive Output Token Generation</div>
+                <div class="block-desc">Generates tokens at constant memory footprint with zero context re-transmission overhead.</div>
               </div>
             </div>
           </div>
 
-          <!-- Mathematical Formulation Card -->
+          <!-- 32-Layer Stack Visual Diagram Card -->
           <div class="content-card" style="margin-top: 1.5rem;">
             <div class="card-head">
-              <h3>📐 Mathematical Physics Formulation</h3>
+              <h3>📐 32-Layer Transformer Stack & O(1) Cache Interception Architecture</h3>
             </div>
-            <div style="font-size: 0.95rem; line-height: 1.8; color: var(--text-secondary);">
-              <p>Let incoming token sequence be represented by embedding vectors $\mathbf{v}_t \in \mathbb{R}^d$ at coordinate $t$. The $O(1)$ continuous state matrix $\Psi \in \mathbb{C}^{B \times d}$ evolves according to:</p>
-              
-              <div class="formula-box">
-                $$\Psi(t) = \Psi(t-1) + \kappa \sum_{b=1}^{B} \Big[ \cos(\omega_b t + \phi_b) + i \sin(\omega_b t + \phi_b) \Big] \mathbf{v}_t$$
+
+            <div style="display: flex; flex-direction: column; gap: 0.8rem;">
+              <div style="background: rgba(0, 240, 255, 0.05); border: 1px solid var(--border-cyan); border-radius: 10px; padding: 0.8rem 1.2rem; display: flex; justify-content: space-between; align-items: center;">
+                <span style="font-weight: 700; font-size: 0.95rem;">STACK: 32 TRANSFORMER HIDDEN LAYERS (Layer 00 – Layer 31)</span>
+                <span class="status-tag tag-pass">ALL 32 LAYERS INTERCEPTED BY KALPANA</span>
+              </div>
+
+              <div style="display: grid; grid-template-columns: repeat(auto-fit, minmax(260px, 1fr)); gap: 0.75rem;">
+                <div style="background: #090d1a; border: 1px solid var(--border); border-radius: 8px; padding: 0.8rem;">
+                  <div style="font-family: var(--font-mono); font-size: 0.8rem; color: var(--cyan); font-weight: 700;">LAYER 00 – 07 (Early Syntax & Tokens)</div>
+                  <div style="font-size: 0.8rem; color: var(--text-muted); margin-top: 0.3rem;">32 Attention Heads ➔ KalpanaCacheLayer (O(1))</div>
+                </div>
+
+                <div style="background: #090d1a; border: 1px solid var(--border); border-radius: 8px; padding: 0.8rem;">
+                  <div style="font-family: var(--font-mono); font-size: 0.8rem; color: var(--cyan); font-weight: 700;">LAYER 08 – 15 (Syntactic & Binding)</div>
+                  <div style="font-size: 0.8rem; color: var(--text-muted); margin-top: 0.3rem;">32 Attention Heads ➔ KalpanaCacheLayer (O(1))</div>
+                </div>
+
+                <div style="background: #090d1a; border: 1px solid var(--border); border-radius: 8px; padding: 0.8rem;">
+                  <div style="font-family: var(--font-mono); font-size: 0.8rem; color: var(--cyan); font-weight: 700;">LAYER 16 – 23 (Semantic Context & Entity)</div>
+                  <div style="font-size: 0.8rem; color: var(--text-muted); margin-top: 0.3rem;">32 Attention Heads ➔ KalpanaCacheLayer (O(1))</div>
+                </div>
+
+                <div style="background: #090d1a; border: 1px solid var(--border); border-radius: 8px; padding: 0.8rem;">
+                  <div style="font-family: var(--font-mono); font-size: 0.8rem; color: var(--cyan); font-weight: 700;">LAYER 24 – 31 (Deep Reasoning & Recall)</div>
+                  <div style="font-size: 0.8rem; color: var(--text-muted); margin-top: 0.3rem;">32 Attention Heads ➔ KalpanaCacheLayer (O(1))</div>
+                </div>
               </div>
 
-              <p>Where:</p>
-              <ul style="padding-left: 1.5rem; margin-top: 0.5rem;">
-                <li>$B = 2,048$: Number of continuous harmonic frequency bands.</li>
-                <li>$\omega_b = \frac{2\pi b}{B}$: Characteristic angular frequency of band $b$.</li>
-                <li>$\phi_b$: Deterministic phase displacement ensuring orthogonality.</li>
-                <li>$\kappa = 1.0$: Coupling constant regulating energy distribution.</li>
-              </ul>
+              <div style="background: #080c18; border-radius: 8px; padding: 1rem; font-size: 0.85rem; color: var(--text-secondary); line-height: 1.6; border: 1px solid var(--border);">
+                <strong>🔒 Proprietary Architecture:</strong> Kalpana RIF replaces the linear KV cache across all layers with proprietary continuous state matrix compilation (International Patent Pending <code>LK/P/1/24089</code>).
+              </div>
             </div>
           </div>
         </div>