Update app.py

app.py

@@ -11,11 +11,22 @@ PKG_PATH = "neuro_semantic_package.pt"
 
 print("🚀 System Startup: Loading Artifacts...")
 if not os.path.exists(PKG_PATH):
-    # Error handling for the web logs
-    raise FileNotFoundError(f"CRITICAL: '{PKG_PATH}' missing. Please upload the .pt file.")
+    # Fallback for local testing if file isn't in root
+    POSSIBLE_PATHS = [
+        "neuro_semantic_package.pt",
+        "/content/drive/MyDrive/Brain2Text_Project/demo_research_v2/neuro_semantic_package.pt"
+    ]
+    for p in POSSIBLE_PATHS:
+        if os.path.exists(p):
+            PKG_PATH = p
+            break
+    
+    if not os.path.exists(PKG_PATH):
+        raise FileNotFoundError(f"CRITICAL: '{PKG_PATH}' missing. Please upload the .pt file.")
 
 # Load the "Black Box" package
-PKG = torch.load(PKG_PATH, map_location="cpu", weights_only=False) # Load to CPU for HF Spaces
+# map_location='cpu' ensures it runs on basic HF spaces without GPU if needed
+PKG = torch.load(PKG_PATH, map_location="cpu", weights_only=False) 
 DATA = PKG['data']
 MODELS = PKG['models']       # The Projectors
 MATRIX = PKG['matrix']       # Pre-calculated Accuracy Table
@@ -111,19 +122,20 @@ def decode_neuro_semantics(subject, projector_alias, text):
         "Match": match_icon
     }])
     
-    return df, f"**Prediction Status:** {match_icon}"
+    return df
 
 def run_batch_analysis(subject, projector_alias):
     # Runs 5 random samples for robust demo
     subject_data = DATA[subject]
     total_indices = list(range(len(subject_data['Text'])))
+    # Sample up to 5 sentences
     selected_indices = random.sample(total_indices, min(5, len(total_indices)))
     
     results = []
     
     for idx in selected_indices:
         txt = subject_data['Text'][idx]
-        df, stat = decode_neuro_semantics(subject, projector_alias, txt)
+        df = decode_neuro_semantics(subject, projector_alias, txt)
         results.append(df)
         
     final_df = pd.concat(results)
@@ -134,61 +146,81 @@ def run_batch_analysis(subject, projector_alias):
 
 # --- 3. UI LAYOUT ---
 
-INTRODUCTION = """
-### 🔬 Abstract & Methodology
-**Goal:** Zero-Shot decoding of emotional sentiment from raw EEG signals.
+# Formatted Report Text
+REPORT_TEXT = """
+### 1. Abstract
+This interface demonstrates a **Brain-Computer Interface (BCI)** capable of decoding high-level semantic information directly from non-invasive EEG signals. By aligning biological neural activity with the latent space of Large Language Models (LLMs), we show that it is possible to reconstruct the **emotional sentiment** of a sentence a user is reading, even if the model has **never seen that user's brain data before**.
+
+### 2. The Dataset: ZuCo (Zurich Cognitive Language Processing Corpus)
+This project utilizes the **ZuCo 2.0 dataset**, a benchmark for cognitive modeling.
+*   **Protocol:** Subjects read movie reviews naturally while their brain activity (EEG) and eye movements were recorded.
+*   **The Challenge:** Unlike synthetic tasks, natural reading involves rapid, complex cognitive processing, making signal decoding significantly harder.
+
+### 3. Methodology: Latent Space Projection
+Instead of training a simple classifier to predict "Positive" or "Negative" from brain waves, we employ a **Neuro-Semantic Projector**.
+*   **The Goal:** To learn a mapping function `f(EEG) → R^768` that transforms raw brain signals into the high-dimensional embedding space of **RoBERTa**.
+*   **The Mechanism:** The system projects the EEG signal into a vector. This vector is then fed into a frozen, pre-trained LLM (`roberta-base-go_emotions`) to generate a probability distribution over **28 distinct emotional states** (e.g., *Admiration, Annoyance, Gratitude, Remorse*).
 
-**Methodology:**
-1.  **Input:** EEG signals from the ZuCo 2.0 dataset (Movie Reviews).
-2.  **Projection:** A Ridge Regression model maps EEG features ($f(EEG)$) to the **RoBERTa-GoEmotions** latent space ($\mathbb{R}^{768}$).
-3.  **Inference:** The projected vector is classified by the frozen RoBERTa head to recover the sentiment probability distribution.
+### 4. Experimental Setup: Strict Zero-Shot Evaluation
+To ensure scientific rigor, this demo adheres to a **Strict Leave-One-Group-Out** protocol.
+*   **Disjoint Training:** The "Projectors" available in this demo were trained on a subset of subjects and validated on **completely different subjects**.
+*   **No Calibration:** The model does not receive any calibration data from the target subject. It must rely on universal neural patterns shared across humans.
 
-**Evaluation Metric:** A prediction is correct if the **Top-1 Brain Prediction** appears within the **Top-2 Text Predictions**.
+### 5. Interpretation of Results
+The demo compares two probability distributions for every sentence:
+1.  **Text Ground Truth:** What the AI model thinks the sentence means based on the text alone.
+2.  **Brain Prediction:** What the AI model thinks the sentence means based **only** on the user's brain waves.
+
+**Accuracy Metric:** A prediction is considered correct if the **Top-1 Emotion** predicted from the Brain Signal matches either the **#1 or #2 Emotion** predicted from the Text.
 """
 
-with gr.Blocks(theme=gr.themes.Soft()) as demo:
+with gr.Blocks(theme=gr.themes.Soft(), title="Neuro-Semantic Decoder") as demo:
     gr.Markdown("# 🧠 Neuro-Semantic Alignment: Zero-Shot Decoding")
     
-    with gr.Accordion("📘 Read Project Report (Abstract & Methodology)", open=False):
-        gr.Markdown(INTRODUCTION)
-    
-    with gr.Row():
-        with gr.Column(scale=1):
-            gr.Markdown("### ⚙️ Configuration")
+    with gr.Tabs():
+        # --- TAB 1: INTERACTIVE DEMO ---
+        with gr.TabItem("🔮 Interactive Demo"):
+            with gr.Row():
+                with gr.Column(scale=1):
+                    gr.Markdown("### ⚙️ Configuration")
+                    
+                    # Selectors
+                    sub_dropdown = gr.Dropdown(choices=list(DATA.keys()), value="ZKB", label="Select Target Subject (Data Source)")
+                    proj_dropdown = gr.Dropdown(choices=list(MODELS.keys()), value="Projector A", label="Select Projector (Decoding Model)")
+                    
+                    # Dynamic Info Boxes
+                    warning_box = gr.Markdown("✅ **VALID ZERO-SHOT CONFIGURATION**\n\nTarget Subject was NOT seen during Projector training.")
+                    history_box = gr.Markdown("**Historical Compatibility:** 40.0%")
+                    
+                    btn = gr.Button("🔮 Run Batch Analysis (5 Samples)", variant="primary")
+                    
+                with gr.Column(scale=2):
+                    gr.Markdown("### 📊 Decoding Results")
+                    
+                    # Output Table
+                    result_table = gr.Dataframe(
+                        headers=["Sentence Stimulus", "Text Ground Truth (Top 2)", "Brain Decoding (Top 3)", "Match"],
+                        wrap=True
+                    )
+                    batch_accuracy_box = gr.Markdown("**Batch Accuracy:** -")
+
+            # Interactivity
+            sub_dropdown.change(fn=get_warning_status, inputs=[sub_dropdown, proj_dropdown], outputs=warning_box)
+            sub_dropdown.change(fn=get_historical_accuracy, inputs=[sub_dropdown, proj_dropdown], outputs=history_box)
             
-            # Selectors
-            sub_dropdown = gr.Dropdown(choices=list(DATA.keys()), value="ZKB", label="Select Target Subject (Data Source)")
-            proj_dropdown = gr.Dropdown(choices=list(MODELS.keys()), value="Projector A", label="Select Projector (Decoding Model)")
+            proj_dropdown.change(fn=get_warning_status, inputs=[sub_dropdown, proj_dropdown], outputs=warning_box)
+            proj_dropdown.change(fn=get_historical_accuracy, inputs=[sub_dropdown, proj_dropdown], outputs=history_box)
             
-            # Dynamic Info Boxes
-            warning_box = gr.Markdown("✅ **VALID ZERO-SHOT CONFIGURATION**\n\nTarget Subject was NOT seen during Projector training.")
-            history_box = gr.Markdown("**Historical Compatibility:** 40.0%")
-            
-            btn = gr.Button("🔮 Run Batch Analysis (5 Samples)", variant="primary")
-            
-        with gr.Column(scale=2):
-            gr.Markdown("### 📊 Decoding Results")
-            
-            # Output Table
-            result_table = gr.Dataframe(
-                headers=["Sentence Stimulus", "Text Ground Truth (Top 2)", "Brain Decoding (Top 3)", "Match"],
-                wrap=True
+            # Run
+            btn.click(
+                fn=run_batch_analysis,
+                inputs=[sub_dropdown, proj_dropdown],
+                outputs=[result_table, batch_accuracy_box]
             )
-            batch_accuracy_box = gr.Markdown("**Batch Accuracy:** -")
 
-    # Interactivity
-    sub_dropdown.change(fn=get_warning_status, inputs=[sub_dropdown, proj_dropdown], outputs=warning_box)
-    sub_dropdown.change(fn=get_historical_accuracy, inputs=[sub_dropdown, proj_dropdown], outputs=history_box)
-    
-    proj_dropdown.change(fn=get_warning_status, inputs=[sub_dropdown, proj_dropdown], outputs=warning_box)
-    proj_dropdown.change(fn=get_historical_accuracy, inputs=[sub_dropdown, proj_dropdown], outputs=history_box)
-    
-    # Run
-    btn.click(
-        fn=run_batch_analysis,
-        inputs=[sub_dropdown, proj_dropdown],
-        outputs=[result_table, batch_accuracy_box]
-    )
+        # --- TAB 2: REPORT ---
+        with gr.TabItem("📘 Project Report"):
+            gr.Markdown(REPORT_TEXT)
 
 if __name__ == "__main__":
     demo.launch()