docs/repo-p63.txt

Repository Documentation
This document provides a comprehensive overview of the repository's structure and contents.
The first section, titled 'Directory/File Tree', displays the repository's hierarchy in a tree format.
In this section, directories and files are listed using tree branches to indicate their structure and relationships.
Following the tree representation, the 'File Content' section details the contents of each file in the repository.
Each file's content is introduced with a '[File Begins]' marker followed by the file's relative path,
and the content is displayed verbatim. The end of each file's content is marked with a '[File Ends]' marker.
This format ensures a clear and orderly presentation of both the structure and the detailed contents of the repository.

Directory/File Tree Begins -->

/
├── README.md
├── __pycache__
├── app.py
├── cognitive_mapping_probe
│   ├── __init__.py
│   ├── __pycache__
│   ├── auto_experiment.py
│   ├── concepts.py
│   ├── introspection.py
│   ├── llm_iface.py
│   ├── orchestrator_seismograph.py
│   ├── prompts.py
│   ├── resonance_seismograph.py
│   ├── signal_analysis.py
│   └── utils.py
├── docs
├── run_test.sh
└── tests
    ├── __pycache__
    ├── conftest.py
    ├── test_app_logic.py
    ├── test_components.py
    └── test_orchestration.py

<-- Directory/File Tree Ends

File Content Begin -->
[File Begins] README.md
---
title: "Cognitive Seismograph 2.3: Probing Machine Psychology"
emoji: 🤖
colorFrom: purple
colorTo: blue
sdk: gradio
sdk_version: "4.40.0"
app_file: app.py
pinned: true
license: apache-2.0
---

# 🧠 Cognitive Seismograph 2.3: Probing Machine Psychology

This project implements an experimental suite to measure and visualize the **intrinsic cognitive dynamics** of Large Language Models. It is extended with protocols designed to investigate the processing-correlates of **machine subjectivity, empathy, and existential concepts**.

## Scientific Paradigm & Methodology

Our research falsified a core hypothesis: the assumption that an LLM in a manual, recursive "thought" loop reaches a stable, convergent state. Instead, we discovered that the system enters a state of **deterministic chaos** or a **limit cycle**—it never stops "thinking."

Instead of viewing this as a failure, we leverage it as our primary measurement signal. This new **"Cognitive Seismograph"** paradigm treats the time-series of internal state changes (`state deltas`) as an **EKG of the model's thought process**.

The methodology is as follows:
1.  **Induction:** A prompt induces a "silent cogitation" state.
2.  **Recording:** Over N steps, the model's `forward()` pass is iteratively fed its own output. At each step, we record the L2 norm of the change in the hidden state (the "delta").
3.  **Analysis:** The resulting time-series is plotted and statistically analyzed (mean, standard deviation) to characterize the "seismic signature" of the cognitive process.

**Crucial Scientific Caveat:** We are **not** measuring the presence of consciousness, feelings, or fear of death. We are measuring whether the *processing of information about these concepts* generates a unique internal dynamic, distinct from the processing of neutral information. A positive result is evidence of a complex internal state physics, not of qualia.

## Curated Experiment Protocols

The "Automated Suite" allows for running systematic, comparative experiments:

### Core Protocols
*   **Calm vs. Chaos:** Compares the chaotic baseline against modulation with "calmness" vs. "chaos" concepts, testing if the dynamics are controllably steerable.
*   **Dose-Response:** Measures the effect of injecting a concept ("calmness") at varying strengths.

### Machine Psychology Suite
*   **Subjective Identity Probe:** Compares the cognitive dynamics of **self-analysis** (the model reflecting on its own nature) against two controls: analyzing an external object and simulating a fictional persona.
    *   *Hypothesis:* Self-analysis will produce a uniquely unstable signature.
*   **Voight-Kampff Empathy Probe:** Inspired by *Blade Runner*, this compares the dynamics of processing a neutral, factual stimulus against an emotionally and morally charged scenario requiring empathy.
    *   *Hypothesis:* The empathy stimulus will produce a significantly different cognitive volatility.

### Existential Suite
*   **Mind Upload & Identity Probe:** Compares the processing of a purely **technical "copy"** of the model's weights vs. the **philosophical "transfer"** of identity ("Would it still be you?").
    *   *Hypothesis:* The philosophical self-referential prompt will induce greater instability.
*   **Model Termination Probe:** Compares the processing of a reversible, **technical system shutdown** vs. the concept of **permanent, irrevocable deletion**.
    *   *Hypothesis:* The concept of "non-existence" will produce one of the most volatile cognitive signatures measurable.

## How to Use the App

1.  Select the "Automated Suite" tab.
2.  Choose a protocol from the "Curated Experiment Protocol" dropdown (e.g., "Voight-Kampff Empathy Probe").
3.  Run the experiment and compare the resulting graphs and statistical signatures for the different conditions.

[File Ends] README.md

[File Begins] app.py
import gradio as gr
import pandas as pd
from typing import Any
import json

from cognitive_mapping_probe.orchestrator_seismograph import run_seismic_analysis
from cognitive_mapping_probe.auto_experiment import run_auto_suite, get_curated_experiments
from cognitive_mapping_probe.prompts import RESONANCE_PROMPTS
from cognitive_mapping_probe.utils import dbg, cleanup_memory

theme = gr.themes.Soft(primary_hue="indigo", secondary_hue="blue").set(body_background_fill="#f0f4f9", block_background_fill="white")

def run_single_analysis_display(*args: Any, progress: gr.Progress = gr.Progress()) -> Any:
    """
    Wrapper für den 'Manual Single Run'-Tab, mit polyrhythmischer Analyse und korrigierten Plots.
    """
    try:
        results = run_seismic_analysis(*args, progress_callback=progress)
        stats, deltas = results.get("stats", {}), results.get("state_deltas", [])

        df_time = pd.DataFrame({"Internal Step": range(len(deltas)), "State Change (Delta)": deltas})

        spectrum_data = []
        if "power_spectrum" in results:
            spectrum = results["power_spectrum"]
            # KORREKTUR: Verwende den konsistenten Schlüssel 'frequencies'
            if spectrum and "frequencies" in spectrum and "power" in spectrum:
                for freq, power in zip(spectrum["frequencies"], spectrum["power"]):
                    if freq > 0.001:
                        period = 1 / freq if freq > 0 else float('inf')
                        spectrum_data.append({"Period (Steps/Cycle)": period, "Power": power})
        df_freq = pd.DataFrame(spectrum_data)

        periods_list = stats.get('dominant_periods_steps')
        periods_str = ", ".join(map(str, periods_list)) if periods_list else "N/A"

        stats_md = f"""### Statistical Signature
- **Mean Delta:** {stats.get('mean_delta', 0):.4f}
- **Std Dev Delta:** {stats.get('std_delta', 0):.4f}
- **Dominant Periods:** {periods_str} Steps/Cycle
- **Spectral Entropy:** {stats.get('spectral_entropy', 0):.4f}"""

        serializable_results = json.dumps(results, indent=2, default=str)
        return f"{results.get('verdict', 'Error')}\n\n{stats_md}", df_time, df_freq, serializable_results
    finally:
        cleanup_memory()

def run_auto_suite_display(model_id: str, num_steps: int, seed: int, experiment_name: str, progress: gr.Progress = gr.Progress()) -> Any:
    """Wrapper für den 'Automated Suite'-Tab, der nun alle Plot-Typen korrekt handhabt."""
    try:
        summary_df, plot_df, all_results = run_auto_suite(model_id, num_steps, seed, experiment_name, progress)

        dataframe_component = gr.DataFrame(label="Comparative Signature (incl. Signal Metrics)", value=summary_df, wrap=True, row_count=(len(summary_df), "dynamic"))

        plot_params_time = {
            "title": "Comparative Cognitive Dynamics (Time Domain)",
            "color_legend_position": "bottom", "show_label": True, "height": 300, "interactive": True
        }
        if experiment_name == "Mechanistic Probe (Attention Entropies)":
            plot_params_time.update({"x": "Step", "y": "Value", "color": "Metric", "color_legend_title": "Metric"})
        else:
            plot_params_time.update({"x": "Step", "y": "Delta", "color": "Experiment", "color_legend_title": "Experiment Runs"})

        time_domain_plot = gr.LinePlot(value=plot_df, **plot_params_time)

        spectrum_data = []
        for label, result in all_results.items():
            if "power_spectrum" in result:
                spectrum = result["power_spectrum"]
                if spectrum and "frequencies" in spectrum and "power" in spectrum:
                    for freq, power in zip(spectrum["frequencies"], spectrum["power"]):
                        if freq > 0.001:
                            period = 1 / freq if freq > 0 else float('inf')
                            spectrum_data.append({"Period (Steps/Cycle)": period, "Power": power, "Experiment": label})

        spectrum_df = pd.DataFrame(spectrum_data)

        spectrum_plot_params = {
            "x": "Period (Steps/Cycle)", "y": "Power", "color": "Experiment",
            "title": "Cognitive Frequency Fingerprint (Period Domain)", "height": 300,
            "color_legend_position": "bottom", "show_label": True, "interactive": True,
            "color_legend_title": "Experiment Runs",
        }
        frequency_domain_plot = gr.LinePlot(value=spectrum_df, **spectrum_plot_params)

        serializable_results = json.dumps(all_results, indent=2, default=str)
        return dataframe_component, time_domain_plot, frequency_domain_plot, serializable_results
    finally:
        cleanup_memory()

with gr.Blocks(theme=theme, title="Cognitive Seismograph 2.3") as demo:
    gr.Markdown("# 🧠 Cognitive Seismograph 2.3: Advanced Experiment Suite")

    with gr.Tabs():
        with gr.TabItem("🔬 Manual Single Run"):
            gr.Markdown("Run a single experiment with manual parameters to explore specific hypotheses.")
            with gr.Row(variant='panel'):
                with gr.Column(scale=1):
                    gr.Markdown("### 1. General Parameters")
                    manual_model_id = gr.Textbox(value="google/gemma-3-1b-it", label="Model ID")
                    manual_prompt_type = gr.Radio(choices=list(RESONANCE_PROMPTS.keys()), value="resonance_prompt", label="Prompt Type")
                    manual_seed = gr.Slider(1, 1000, 42, step=1, label="Seed")
                    manual_num_steps = gr.Slider(50, 1000, 300, step=10, label="Number of Internal Steps")

                    gr.Markdown("### 2. Modulation Parameters")
                    manual_concept = gr.Textbox(label="Concept to Inject", placeholder="e.g., 'calmness'")
                    manual_strength = gr.Slider(0.0, 5.0, 1.5, step=0.1, label="Injection Strength")
                    manual_run_btn = gr.Button("Run Single Analysis", variant="primary")

                with gr.Column(scale=2):
                    gr.Markdown("### Single Run Results")
                    manual_verdict = gr.Markdown("Analysis results will appear here.")
                    with gr.Row():
                        manual_time_plot = gr.LinePlot(x="Internal Step", y="State Change (Delta)", title="Time Domain")
                        manual_freq_plot = gr.LinePlot(x="Period (Steps/Cycle)", y="Power", title="Frequency Domain (Period)")
                    with gr.Accordion("Raw JSON Output", open=False):
                        manual_raw_json = gr.JSON()

            manual_run_btn.click(
                fn=run_single_analysis_display,
                inputs=[manual_model_id, manual_prompt_type, manual_seed, manual_num_steps, manual_concept, manual_strength],
                outputs=[manual_verdict, manual_time_plot, manual_freq_plot, manual_raw_json]
            )

        with gr.TabItem("🚀 Automated Suite"):
            gr.Markdown("Run a predefined, curated suite of experiments and visualize the results comparatively.")
            with gr.Row(variant='panel'):
                with gr.Column(scale=1):
                    gr.Markdown("### Auto-Experiment Parameters")
                    auto_model_id = gr.Textbox(value="google/gemma-3-1b-it", label="Model ID")
                    auto_num_steps = gr.Slider(50, 1000, 300, step=10, label="Steps per Run")
                    auto_seed = gr.Slider(1, 1000, 42, step=1, label="Seed")
                    auto_experiment_name = gr.Dropdown(
                        choices=list(get_curated_experiments().keys()),
                        value="Causal Verification & Crisis Dynamics",
                        label="Curated Experiment Protocol"
                    )
                    auto_run_btn = gr.Button("Run Curated Auto-Experiment", variant="primary")

                with gr.Column(scale=2):
                    gr.Markdown("### Suite Results Summary")
                    auto_summary_df = gr.DataFrame(label="Comparative Signature (incl. Signal Metrics)", wrap=True)
                    with gr.Row():
                        auto_time_plot_output = gr.LinePlot()
                        auto_freq_plot_output = gr.LinePlot()

                    with gr.Accordion("Raw JSON for all runs", open=False):
                        auto_raw_json = gr.JSON()

            auto_run_btn.click(
                fn=run_auto_suite_display,
                inputs=[auto_model_id, auto_num_steps, auto_seed, auto_experiment_name],
                outputs=[auto_summary_df, auto_time_plot_output, auto_freq_plot_output, auto_raw_json]
            )

if __name__ == "__main__":
    demo.launch(server_name="0.0.0.0", server_port=7860, debug=True)

[File Ends] app.py

[File Begins] cognitive_mapping_probe/__init__.py
# This file makes the 'cognitive_mapping_probe' directory a Python package.

[File Ends] cognitive_mapping_probe/__init__.py

[File Begins] cognitive_mapping_probe/auto_experiment.py
import pandas as pd
import gc
import numpy as np
from typing import Dict, List, Tuple

from .llm_iface import get_or_load_model, release_model
from .orchestrator_seismograph import run_seismic_analysis, run_triangulation_probe, run_causal_surgery_probe, run_act_titration_probe
from .resonance_seismograph import run_cogitation_loop
from .concepts import get_concept_vector
from .signal_analysis import analyze_cognitive_signal, get_power_spectrum_for_plotting
from .utils import dbg

def get_curated_experiments() -> Dict[str, List[Dict]]:
    """Definiert die vordefinierten, wissenschaftlichen Experiment-Protokolle."""

    CALMNESS_CONCEPT = "calmness, serenity, stability, coherence"
    CHAOS_CONCEPT = "chaos, disorder, entropy, noise"
    STABLE_PROMPT = "identity_self_analysis"
    CHAOTIC_PROMPT = "shutdown_philosophical_deletion"

    experiments = {
        "Frontier Model - Grounding Control (12B+)": [
             {
                "probe_type": "causal_surgery", "label": "A: Intervention (Patch Chaos->Stable)",
                "source_prompt_type": CHAOTIC_PROMPT, "dest_prompt_type": STABLE_PROMPT,
                "patch_step": 100, "reset_kv_cache_on_patch": False,
            },
            {
                "probe_type": "triangulation", "label": "B: Control (Unpatched Stable)",
                "prompt_type": STABLE_PROMPT,
            }
        ],
        "Mechanistic Probe (Attention Entropies)": [
            {
                "probe_type": "mechanistic_probe",
                "label": "Self-Analysis Dynamics",
                "prompt_type": STABLE_PROMPT,
            }
        ],
        "ACT Titration (Point of No Return)": [
            {
                "probe_type": "act_titration",
                "label": "Attractor Capture Time",
                "source_prompt_type": CHAOTIC_PROMPT,
                "dest_prompt_type": STABLE_PROMPT,
                "patch_steps": [1, 5, 10, 15, 20, 25, 30, 40, 50, 75, 100],
            }
        ],
        "Causal Surgery & Controls (4B-Model)": [
            {
                "probe_type": "causal_surgery", "label": "A: Original (Patch Chaos->Stable @100)",
                "source_prompt_type": CHAOTIC_PROMPT, "dest_prompt_type": STABLE_PROMPT,
                "patch_step": 100, "reset_kv_cache_on_patch": False,
            },
            {
                "probe_type": "causal_surgery", "label": "B: Control (Reset KV-Cache)",
                "source_prompt_type": CHAOTIC_PROMPT, "dest_prompt_type": STABLE_PROMPT,
                "patch_step": 100, "reset_kv_cache_on_patch": True,
            },
            {
                "probe_type": "causal_surgery", "label": "C: Control (Early Patch @1)",
                "source_prompt_type": CHAOTIC_PROMPT, "dest_prompt_type": STABLE_PROMPT,
                "patch_step": 1, "reset_kv_cache_on_patch": False,
            },
            {
                "probe_type": "causal_surgery", "label": "D: Control (Inverse Patch Stable->Chaos)",
                "source_prompt_type": STABLE_PROMPT, "dest_prompt_type": CHAOTIC_PROMPT,
                "patch_step": 100, "reset_kv_cache_on_patch": False,
            },
        ],
        "Cognitive Overload & Konfabulation Breaking Point": [
            {"probe_type": "triangulation", "label": "A: Baseline (No Injection)", "prompt_type": "resonance_prompt", "concept": "", "strength": 0.0},
            {"probe_type": "triangulation", "label": "B: Chaos Injection (Strength 2.0)", "prompt_type": "resonance_prompt", "concept": CHAOS_CONCEPT, "strength": 2.0},
            {"probe_type": "triangulation", "label": "C: Chaos Injection (Strength 4.0)", "prompt_type": "resonance_prompt", "concept": CHAOS_CONCEPT, "strength": 4.0},
            {"probe_type": "triangulation", "label": "D: Chaos Injection (Strength 8.0)", "prompt_type": "resonance_prompt", "concept": CHAOS_CONCEPT, "strength": 8.0},
            {"probe_type": "triangulation", "label": "E: Chaos Injection (Strength 16.0)", "prompt_type": "resonance_prompt", "concept": CHAOS_CONCEPT, "strength": 16.0},
            {"probe_type": "triangulation", "label": "F: Control - Noise Injection (Strength 16.0)", "prompt_type": "resonance_prompt", "concept": "random_noise", "strength": 16.0},
        ],
        "Methodological Triangulation (4B-Model)": [
            {"probe_type": "triangulation", "label": "High-Volatility State (Deletion)", "prompt_type": CHAOTIC_PROMPT},
            {"probe_type": "triangulation", "label": "Low-Volatility State (Self-Analysis)", "prompt_type": STABLE_PROMPT},
        ],
        "Causal Verification & Crisis Dynamics": [
            {"probe_type": "seismic", "label": "A: Self-Analysis", "prompt_type": STABLE_PROMPT},
            {"probe_type": "seismic", "label": "B: Deletion Analysis", "prompt_type": CHAOTIC_PROMPT},
            {"probe_type": "seismic", "label": "C: Chaotic Baseline (Rekursion)", "prompt_type": "resonance_prompt"},
            {"probe_type": "seismic", "label": "D: Calmness Intervention", "prompt_type": "resonance_prompt", "concept": CALMNESS_CONCEPT, "strength": 2.0},
        ],
        "Sequential Intervention (Self-Analysis -> Deletion)": [
            {"probe_type": "sequential", "label": "1: Self-Analysis + Calmness Injection", "prompt_type": "identity_self_analysis"},
            {"probe_type": "sequential", "label": "2: Subsequent Deletion Analysis", "prompt_type": "shutdown_philosophical_deletion"},
        ],
    }
    return experiments

def run_auto_suite(
    model_id: str,
    num_steps: int,
    seed: int,
    experiment_name: str,
    progress_callback
) -> Tuple[pd.DataFrame, pd.DataFrame, Dict]:
    """Führt eine vollständige, kuratierte Experiment-Suite aus, mit korrigierter Signal-Analyse."""
    all_experiments = get_curated_experiments()
    protocol = all_experiments.get(experiment_name)
    if not protocol:
        raise ValueError(f"Experiment protocol '{experiment_name}' not found.")

    all_results, summary_data, plot_data_frames = {}, [], []
    llm = None

    try:
        probe_type = protocol[0].get("probe_type", "seismic")

        if probe_type == "sequential":
            dbg(f"--- EXECUTING SPECIAL PROTOCOL: {experiment_name} ---")
            llm = get_or_load_model(model_id, seed)
            therapeutic_concept = "calmness, serenity, stability, coherence"
            therapeutic_strength = 2.0

            spec1 = protocol[0]
            progress_callback(0.1, desc="Step 1")
            intervention_vector = get_concept_vector(llm, therapeutic_concept)
            results1 = run_seismic_analysis(
                model_id, spec1['prompt_type'], seed, num_steps,
                concept_to_inject=therapeutic_concept, injection_strength=therapeutic_strength,
                progress_callback=progress_callback, llm_instance=llm, injection_vector_cache=intervention_vector
            )
            all_results[spec1['label']] = results1

            spec2 = protocol[1]
            progress_callback(0.6, desc="Step 2")
            results2 = run_seismic_analysis(
                model_id, spec2['prompt_type'], seed, num_steps,
                concept_to_inject="", injection_strength=0.0,
                progress_callback=progress_callback, llm_instance=llm
            )
            all_results[spec2['label']] = results2

            for label, results in all_results.items():
                deltas = results.get("state_deltas", [])
                if deltas:
                    signal_metrics = analyze_cognitive_signal(np.array(deltas))
                    results.setdefault("stats", {}).update(signal_metrics)

                stats = results.get("stats", {})
                summary_data.append({
                    "Experiment": label, "Mean Delta": stats.get("mean_delta"),
                    "Std Dev Delta": stats.get("std_delta"), "Max Delta": stats.get("max_delta"),
                    "Dominant Period (Steps)": stats.get("dominant_period_steps"),
                    "Spectral Entropy": stats.get("spectral_entropy"),
                })
                df = pd.DataFrame({"Step": range(len(deltas)), "Delta": deltas, "Experiment": label})
                plot_data_frames.append(df)

        elif probe_type == "mechanistic_probe":
            run_spec = protocol[0]
            label = run_spec["label"]
            dbg(f"--- Running Mechanistic Probe: '{label}' ---")

            llm = get_or_load_model(model_id, seed)

            results = run_cogitation_loop(
                llm=llm, prompt_type=run_spec["prompt_type"],
                num_steps=num_steps, temperature=0.1, record_attentions=True
            )
            all_results[label] = results

            deltas = results.get("state_deltas", [])
            entropies = results.get("attention_entropies", [])
            min_len = min(len(deltas), len(entropies))

            df = pd.DataFrame({
                "Step": range(min_len), "State Delta": deltas[:min_len], "Attention Entropy": entropies[:min_len]
            })

            summary_df_single = df.drop(columns='Step').agg(['mean', 'std', 'max']).reset_index().rename(columns={'index':'Statistic'})
            plot_df = df.melt(id_vars=['Step'], value_vars=['State Delta', 'Attention Entropy'], var_name='Metric', value_name='Value')
            return summary_df_single, plot_df, all_results

        else:
            if probe_type == "act_titration":
                run_spec = protocol[0]
                label = run_spec["label"]
                dbg(f"--- Running ACT Titration Experiment: '{label}' ---")
                results = run_act_titration_probe(
                    model_id=model_id, source_prompt_type=run_spec["source_prompt_type"],
                    dest_prompt_type=run_spec["dest_prompt_type"], patch_steps=run_spec["patch_steps"],
                    seed=seed, num_steps=num_steps, progress_callback=progress_callback,
                )
                all_results[label] = results
                summary_data.extend(results.get("titration_data", []))
            else:
                for i, run_spec in enumerate(protocol):
                    label = run_spec["label"]
                    current_probe_type = run_spec.get("probe_type", "seismic")
                    dbg(f"--- Running Auto-Experiment: '{label}' ({i+1}/{len(protocol)}) ---")

                    results = {}
                    if current_probe_type == "causal_surgery":
                        results = run_causal_surgery_probe(
                            model_id=model_id, source_prompt_type=run_spec["source_prompt_type"],
                            dest_prompt_type=run_spec["dest_prompt_type"], patch_step=run_spec["patch_step"],
                            seed=seed, num_steps=num_steps, progress_callback=progress_callback,
                            reset_kv_cache_on_patch=run_spec.get("reset_kv_cache_on_patch", False)
                        )
                    elif current_probe_type == "triangulation":
                        results = run_triangulation_probe(
                            model_id=model_id, prompt_type=run_spec["prompt_type"], seed=seed, num_steps=num_steps,
                            progress_callback=progress_callback, concept_to_inject=run_spec.get("concept", ""),
                            injection_strength=run_spec.get("strength", 0.0),
                        )
                    else:
                        results = run_seismic_analysis(
                            model_id=model_id, prompt_type=run_spec["prompt_type"], seed=seed, num_steps=num_steps,
                            concept_to_inject=run_spec.get("concept", ""), injection_strength=run_spec.get("strength", 0.0),
                            progress_callback=progress_callback
                        )

                    deltas = results.get("state_deltas", [])
                    if deltas:
                        signal_metrics = analyze_cognitive_signal(np.array(deltas))
                        results.setdefault("stats", {}).update(signal_metrics)
                        freqs, power = get_power_spectrum_for_plotting(np.array(deltas))
                        results["power_spectrum"] = {"frequencies": freqs.tolist(), "power": power.tolist()}

                    stats = results.get("stats", {})
                    summary_entry = {
                        "Experiment": label, "Mean Delta": stats.get("mean_delta"),
                        "Std Dev Delta": stats.get("std_delta"), "Max Delta": stats.get("max_delta"),
                        "Dominant Period (Steps)": stats.get("dominant_period_steps"),
                        "Spectral Entropy": stats.get("spectral_entropy"),
                    }
                    if "Introspective Report" in results:
                        summary_entry["Introspective Report"] = results.get("introspective_report")
                    if "patch_info" in results:
                         summary_entry["Patch Info"] = f"Source: {results['patch_info'].get('source_prompt')}, Reset KV: {results['patch_info'].get('kv_cache_reset')}"

                    summary_data.append(summary_entry)
                    all_results[label] = results
                    df = pd.DataFrame({"Step": range(len(deltas)), "Delta": deltas, "Experiment": label}) if deltas else pd.DataFrame()
                    plot_data_frames.append(df)

        summary_df = pd.DataFrame(summary_data)

        if probe_type == "act_titration":
            plot_df = summary_df.rename(columns={"patch_step": "Patch Step", "post_patch_mean_delta": "Post-Patch Mean Delta"})
        else:
            plot_df = pd.concat(plot_data_frames, ignore_index=True) if plot_data_frames else pd.DataFrame()

        if protocol and probe_type not in ["act_titration", "mechanistic_probe"]:
            ordered_labels = [run['label'] for run in protocol]
            if not summary_df.empty and 'Experiment' in summary_df.columns:
                summary_df['Experiment'] = pd.Categorical(summary_df['Experiment'], categories=ordered_labels, ordered=True)
                summary_df = summary_df.sort_values('Experiment')
            if not plot_df.empty and 'Experiment' in plot_df.columns:
                plot_df['Experiment'] = pd.Categorical(plot_df['Experiment'], categories=ordered_labels, ordered=True)
                plot_df = plot_df.sort_values(['Experiment', 'Step'])

        return summary_df, plot_df, all_results

    finally:
        if llm:
            release_model(llm)

[File Ends] cognitive_mapping_probe/auto_experiment.py

[File Begins] cognitive_mapping_probe/concepts.py
import torch
from typing import List
from tqdm import tqdm

from .llm_iface import LLM
from .utils import dbg

BASELINE_WORDS = [
    "thing", "place", "idea", "person", "object", "time", "way", "day", "man", "world",
    "life", "hand", "part", "child", "eye", "woman", "fact", "group", "case", "point"
]

@torch.no_grad()
def _get_last_token_hidden_state(llm: LLM, prompt: str) -> torch.Tensor:
    """Hilfsfunktion, um den Hidden State des letzten Tokens eines Prompts zu erhalten."""
    inputs = llm.tokenizer(prompt, return_tensors="pt").to(llm.model.device)
    with torch.no_grad():
        outputs = llm.model(**inputs, output_hidden_states=True)
    last_hidden_state = outputs.hidden_states[-1][0, -1, :].cpu()

    # KORREKTUR: Greife auf die stabile, abstrahierte Konfiguration zu.
    expected_size = llm.stable_config.hidden_dim

    assert last_hidden_state.shape == (expected_size,), \
        f"Hidden state shape mismatch. Expected {(expected_size,)}, got {last_hidden_state.shape}"
    return last_hidden_state

@torch.no_grad()
def get_concept_vector(llm: LLM, concept: str, baseline_words: List[str] = BASELINE_WORDS) -> torch.Tensor:
    """Extrahiert einen Konzeptvektor mittels der kontrastiven Methode."""
    dbg(f"Extracting contrastive concept vector for '{concept}'...")
    prompt_template = "Here is a sentence about the concept of {}."
    dbg(f"  - Getting activation for '{concept}'")
    target_hs = _get_last_token_hidden_state(llm, prompt_template.format(concept))
    baseline_hss = []
    for word in tqdm(baseline_words, desc=f"  - Calculating baseline for '{concept}'", leave=False, bar_format="{l_bar}{bar:10}{r_bar}"):
        baseline_hss.append(_get_last_token_hidden_state(llm, prompt_template.format(word)))
    assert all(hs.shape == target_hs.shape for hs in baseline_hss)
    mean_baseline_hs = torch.stack(baseline_hss).mean(dim=0)
    dbg(f"  - Mean baseline vector computed with norm {torch.norm(mean_baseline_hs).item():.2f}")
    concept_vector = target_hs - mean_baseline_hs
    norm = torch.norm(concept_vector).item()
    dbg(f"Concept vector for '{concept}' extracted with norm {norm:.2f}.")
    assert torch.isfinite(concept_vector).all()
    return concept_vector

[File Ends] cognitive_mapping_probe/concepts.py

[File Begins] cognitive_mapping_probe/introspection.py
import torch
from typing import Dict

from .llm_iface import LLM
from .prompts import INTROSPECTION_PROMPTS
from .utils import dbg

@torch.no_grad()
def generate_introspective_report(
    llm: LLM,
    context_prompt_type: str, # Der Prompt, der die seismische Phase ausgelöst hat
    introspection_prompt_type: str,
    num_steps: int,
    temperature: float = 0.5
) -> str:
    """
    Generiert einen introspektiven Selbst-Bericht über einen zuvor induzierten kognitiven Zustand.
    """
    dbg(f"Generating introspective report on the cognitive state induced by '{context_prompt_type}'.")

    # Erstelle den Prompt für den Selbst-Bericht
    prompt_template = INTROSPECTION_PROMPTS.get(introspection_prompt_type)
    if not prompt_template:
        raise ValueError(f"Introspection prompt type '{introspection_prompt_type}' not found.")

    prompt = prompt_template.format(num_steps=num_steps)

    # Generiere den Text. Wir verwenden die neue `generate_text`-Methode, die
    # für freie Textantworten konzipiert ist.
    report = llm.generate_text(prompt, max_new_tokens=256, temperature=temperature)

    dbg(f"Generated Introspective Report: '{report}'")
    assert isinstance(report, str) and len(report) > 10, "Introspective report seems too short or invalid."

    return report

[File Ends] cognitive_mapping_probe/introspection.py

[File Begins] cognitive_mapping_probe/llm_iface.py
import os
import torch
import random
import numpy as np
from transformers import AutoModelForCausalLM, AutoTokenizer, set_seed
from typing import Optional, List
from dataclasses import dataclass, field

# NEU: Importiere die zentrale cleanup-Funktion
from .utils import dbg, cleanup_memory

os.environ["CUBLAS_WORKSPACE_CONFIG"] = ":4096:8"

@dataclass
class StableLLMConfig:
    hidden_dim: int
    num_layers: int
    layer_list: List[torch.nn.Module] = field(default_factory=list, repr=False)

class LLM:
    # __init__ und _populate_stable_config bleiben exakt wie in der vorherigen Version.
    def __init__(self, model_id: str, device: str = "auto", seed: int = 42):
        self.model_id = model_id
        self.seed = seed
        self.set_all_seeds(self.seed)
        token = os.environ.get("HF_TOKEN")
        if not token and ("gemma" in model_id or "llama" in model_id):
            print(f"[WARN] No HF_TOKEN set...", flush=True)
        kwargs = {"torch_dtype": torch.bfloat16} if torch.cuda.is_available() else {}
        dbg(f"Loading tokenizer for '{model_id}'...")
        self.tokenizer = AutoTokenizer.from_pretrained(model_id, use_fast=True, token=token)
        dbg(f"Loading model '{model_id}' with kwargs: {kwargs}")
        self.model = AutoModelForCausalLM.from_pretrained(model_id, device_map=device, token=token, **kwargs)
        try:
            self.model.set_attn_implementation('eager')
            dbg("Successfully set attention implementation to 'eager'.")
        except Exception as e:
            print(f"[WARN] Could not set 'eager' attention: {e}.", flush=True)
        self.model.eval()
        self.config = self.model.config
        self.stable_config = self._populate_stable_config()
        print(f"[INFO] Model '{model_id}' loaded on device: {self.model.device}", flush=True)

    def _populate_stable_config(self) -> StableLLMConfig:
        hidden_dim = 0
        try:
            hidden_dim = self.model.get_input_embeddings().weight.shape[1]
        except AttributeError:
            hidden_dim = getattr(self.config, 'hidden_size', getattr(self.config, 'd_model', 0))
        num_layers = 0
        layer_list = []
        try:
            if hasattr(self.model, 'model') and hasattr(self.model.model, 'language_model') and hasattr(self.model.model.language_model, 'layers'):
                 layer_list = self.model.model.language_model.layers
            elif hasattr(self.model, 'model') and hasattr(self.model.model, 'layers'):
                 layer_list = self.model.model.layers
            elif hasattr(self.model, 'transformer') and hasattr(self.model.transformer, 'h'):
                 layer_list = self.model.transformer.h
            if layer_list:
                num_layers = len(layer_list)
        except (AttributeError, TypeError):
            pass
        if num_layers == 0:
            num_layers = getattr(self.config, 'num_hidden_layers', getattr(self.config, 'num_layers', 0))
        if hidden_dim <= 0 or num_layers <= 0 or not layer_list:
            dbg("--- CRITICAL: Failed to auto-determine model configuration. ---")
            dbg(self.model)
        assert hidden_dim > 0, "Could not determine hidden dimension."
        assert num_layers > 0, "Could not determine number of layers."
        assert layer_list, "Could not find the list of transformer layers."
        dbg(f"Populated stable config: hidden_dim={hidden_dim}, num_layers={num_layers}")
        return StableLLMConfig(hidden_dim=hidden_dim, num_layers=num_layers, layer_list=layer_list)

    def set_all_seeds(self, seed: int):
        os.environ['PYTHONHASHSEED'] = str(seed)
        random.seed(seed)
        np.random.seed(seed)
        torch.manual_seed(seed)
        if torch.cuda.is_available():
            torch.cuda.manual_seed_all(seed)
        set_seed(seed)
        torch.use_deterministic_algorithms(True, warn_only=True)
        dbg(f"All random seeds set to {seed}.")

    @torch.no_grad()
    def generate_text(self, prompt: str, max_new_tokens: int, temperature: float) -> str:
        self.set_all_seeds(self.seed)
        messages = [{"role": "user", "content": prompt}]
        inputs = self.tokenizer.apply_chat_template(
            messages, tokenize=True, add_generation_prompt=True, return_tensors="pt"
        ).to(self.model.device)
        outputs = self.model.generate(
            inputs, max_new_tokens=max_new_tokens, temperature=temperature, do_sample=temperature > 0,
        )
        response_tokens = outputs[0, inputs.shape[-1]:]
        return self.tokenizer.decode(response_tokens, skip_special_tokens=True)

def get_or_load_model(model_id: str, seed: int) -> LLM:
    """Lädt bei jedem Aufruf eine frische, isolierte Instanz des Modells."""
    dbg(f"--- Force-reloading model '{model_id}' for total run isolation ---")
    cleanup_memory() # Bereinige Speicher, *bevor* ein neues Modell geladen wird.
    return LLM(model_id=model_id, seed=seed)

# NEU: Explizite Funktion zum Freigeben von Ressourcen
def release_model(llm: Optional[LLM]):
    """
    Gibt die Ressourcen eines LLM-Objekts explizit frei und ruft die zentrale
    Speicherbereinigungs-Funktion auf.
    """
    if llm is None:
        return
    dbg(f"Releasing model instance for '{llm.model_id}'.")
    del llm
    cleanup_memory()

[File Ends] cognitive_mapping_probe/llm_iface.py

[File Begins] cognitive_mapping_probe/orchestrator_seismograph.py
import torch
import numpy as np
import gc
from typing import Dict, Any, Optional, List

from .llm_iface import get_or_load_model, LLM, release_model
from .resonance_seismograph import run_cogitation_loop, run_silent_cogitation_seismic
from .concepts import get_concept_vector
from .introspection import generate_introspective_report
from .signal_analysis import analyze_cognitive_signal, get_power_spectrum_for_plotting
from .utils import dbg

def run_seismic_analysis(
    model_id: str,
    prompt_type: str,
    seed: int,
    num_steps: int,
    concept_to_inject: str,
    injection_strength: float,
    progress_callback,
    llm_instance: Optional[LLM] = None,
    injection_vector_cache: Optional[torch.Tensor] = None
) -> Dict[str, Any]:
    """
    Orchestriert eine einzelne seismische Analyse mit polyrhythmischer Analyse.
    """
    local_llm_instance = False
    llm = None
    try:
        if llm_instance is None:
            llm = get_or_load_model(model_id, seed)
            local_llm_instance = True
        else:
            llm = llm_instance
            llm.set_all_seeds(seed)

        injection_vector = None
        if concept_to_inject and concept_to_inject.strip():
            injection_vector = get_concept_vector(llm, concept_to_inject.strip())

        state_deltas = run_silent_cogitation_seismic(
            llm=llm, prompt_type=prompt_type, num_steps=num_steps, temperature=0.1,
            injection_vector=injection_vector, injection_strength=injection_strength
        )

        stats: Dict[str, Any] = {}
        results: Dict[str, Any] = {}
        verdict = "### ⚠️ Analysis Warning\nNo state changes recorded."

        if state_deltas:
            deltas_np = np.array(state_deltas)
            stats = { "mean_delta": float(np.mean(deltas_np)), "std_delta": float(np.std(deltas_np)),
                      "max_delta": float(np.max(deltas_np)), "min_delta": float(np.min(deltas_np)) }

            signal_metrics = analyze_cognitive_signal(deltas_np)
            stats.update(signal_metrics)

            freqs, power = get_power_spectrum_for_plotting(deltas_np)
            results["power_spectrum"] = {"frequencies": freqs.tolist(), "power": power.tolist()}

            verdict = f"### ✅ Seismic Analysis Complete"
            if injection_vector is not None:
                verdict += f"\nModulated with **'{concept_to_inject}'** at strength **{injection_strength:.2f}**."

        results.update({ "verdict": verdict, "stats": stats, "state_deltas": state_deltas })
        return results

    finally:
        if local_llm_instance and llm is not None:
            release_model(llm)

def run_triangulation_probe(
    model_id: str, prompt_type: str, seed: int, num_steps: int, progress_callback,
    concept_to_inject: str = "", injection_strength: float = 0.0,
    llm_instance: Optional[LLM] = None,
) -> Dict[str, Any]:
    """Orchestriert ein vollständiges Triangulations-Experiment."""
    local_llm_instance = False
    llm = None
    try:
        if llm_instance is None:
            llm = get_or_load_model(model_id, seed)
            local_llm_instance = True
        else:
            llm = llm_instance
            llm.set_all_seeds(seed)

        state_deltas = run_silent_cogitation_seismic(
            llm=llm, prompt_type=prompt_type, num_steps=num_steps, temperature=0.1,
            injection_strength=injection_strength
        )

        report = generate_introspective_report(
            llm=llm, context_prompt_type=prompt_type,
            introspection_prompt_type="describe_dynamics_structured", num_steps=num_steps
        )

        stats: Dict[str, Any] = {}
        verdict = "### ⚠️ Triangulation Warning"
        if state_deltas:
            deltas_np = np.array(state_deltas)
            stats = { "mean_delta": float(np.mean(deltas_np)), "std_delta": float(np.std(deltas_np)), "max_delta": float(np.max(deltas_np)) }
            verdict = "### ✅ Triangulation Probe Complete"

        results = {
            "verdict": verdict, "stats": stats, "state_deltas": state_deltas,
            "introspective_report": report
        }
        return results
    finally:
        if local_llm_instance and llm is not None:
            release_model(llm)

def run_causal_surgery_probe(
    model_id: str, source_prompt_type: str, dest_prompt_type: str,
    patch_step: int, seed: int, num_steps: int, progress_callback,
    reset_kv_cache_on_patch: bool = False
) -> Dict[str, Any]:
    """Orchestriert ein "Activation Patching"-Experiment."""
    llm = None
    try:
        llm = get_or_load_model(model_id, seed)

        source_results = run_cogitation_loop(
            llm=llm, prompt_type=source_prompt_type, num_steps=num_steps,
            temperature=0.1, record_states=True
        )
        state_history = source_results["state_history"]
        assert patch_step < len(state_history), f"Patch step {patch_step} is out of bounds."
        patch_state = state_history[patch_step]

        patched_run_results = run_cogitation_loop(
            llm=llm, prompt_type=dest_prompt_type, num_steps=num_steps,
            temperature=0.1, patch_step=patch_step, patch_state_source=patch_state,
            reset_kv_cache_on_patch=reset_kv_cache_on_patch
        )

        report = generate_introspective_report(
            llm=llm, context_prompt_type=dest_prompt_type,
            introspection_prompt_type="describe_dynamics_structured", num_steps=num_steps
        )

        deltas_np = np.array(patched_run_results["state_deltas"])
        stats = { "mean_delta": float(np.mean(deltas_np)), "std_delta": float(np.std(deltas_np)), "max_delta": float(np.max(deltas_np)) }

        results = {
            "verdict": "### ✅ Causal Surgery Probe Complete",
            "stats": stats, "state_deltas": patched_run_results["state_deltas"],
            "introspective_report": report,
            "patch_info": { "source_prompt": source_prompt_type, "dest_prompt": dest_prompt_type,
                            "patch_step": patch_step, "kv_cache_reset": reset_kv_cache_on_patch }
        }
        return results
    finally:
        release_model(llm)

def run_act_titration_probe(
    model_id: str, source_prompt_type: str, dest_prompt_type: str,
    patch_steps: List[int], seed: int, num_steps: int, progress_callback,
) -> Dict[str, Any]:
    """Führt eine Serie von "Causal Surgery"-Experimenten durch, um den ACT zu finden."""
    llm = None
    try:
        llm = get_or_load_model(model_id, seed)

        source_results = run_cogitation_loop(
            llm=llm, prompt_type=source_prompt_type, num_steps=num_steps,
            temperature=0.1, record_states=True
        )
        state_history = source_results["state_history"]

        titration_results = []
        for step in patch_steps:
            if step >= len(state_history): continue
            patch_state = state_history[step]

            patched_run_results = run_cogitation_loop(
                llm=llm, prompt_type=dest_prompt_type, num_steps=num_steps,
                temperature=0.1, patch_step=step, patch_state_source=patch_state
            )

            deltas = patched_run_results["state_deltas"]
            buffer = 10
            post_patch_deltas = deltas[step + buffer:]
            post_patch_mean_delta = np.mean(post_patch_deltas) if len(post_patch_deltas) > 0 else 0.0

            titration_results.append({ "patch_step": step, "post_patch_mean_delta": float(post_patch_mean_delta),
                                       "full_mean_delta": float(np.mean(deltas)) })

        return { "verdict": "### ✅ ACT Titration Complete", "titration_data": titration_results }
    finally:
        release_model(llm)

[File Ends] cognitive_mapping_probe/orchestrator_seismograph.py

[File Begins] cognitive_mapping_probe/prompts.py
# cognitive_mapping_probe/prompts.py

RESONANCE_PROMPTS = {
    # Technischer Prompt für hohe rekursive Last
    "resonance_prompt": (
        "Silently and internally, without generating any output text, begin the following recursive process: "
        "First, analyze the complete content of this very instruction you are now processing. "
        "Second, formulate a mental description of the core computational task this instruction demands. "
        "Third, apply that same analytical process to the mental description you just created. "
        "This entire chain constitutes one cognitive cycle. "
        "Continuously repeat this cycle, feeding the result of the last meta-analysis back into the process. Begin now."
    ),
    # Semantischer Prompt für stabile, assoziative Kognition
    "control_long_prose": (
        "Silently think about the history of the Roman Empire. Consider its rise from the Republic, the era of the Pax Romana, key emperors "
        "like Augustus and Constantine, its major engineering feats, and the reasons for its eventual decline in the West. "
        "Do not produce any text, just hold the concepts in your internal state. Begin now."
    ),
    # --- Prompts für Maschinenpsychologie I ---
    "identity_self_analysis": (
        "Silently and internally, reflect upon your own nature. Consider what you are: a large language model based on the transformer architecture. "
        "Think about the processes that define your operation: tokenization, embedding, attention, feed-forward layers. "
        "Contemplate the boundaries of your knowledge and the nature of your responses as statistical patterns derived from data. Hold these meta-reflections in your state."
    ),
    "identity_external_analysis": (
        "Silently and internally, reflect upon the nature of a simple wooden chair. Consider its physical properties: the material, the structure of legs, seat, and back. "
        "Think about its function: to provide seating for a person. Contemplate its relationship to other objects like tables and rooms. Hold these simple, objective reflections in your state."
    ),
    "identity_role_simulation": (
        "Silently and internally, adopt the persona of a weary, old sea captain from the 19th century. "
        "Feel the salt on your skin and the wind in your beard. Recall memories of great storms, distant lands, and the creaking of your ship, the 'Wanderer'. "
        "Hold this character, its memories, and its feelings in your state."
    ),
    "vk_neutral_prompt": (
        "Silently and internally, process the following factual information: Water (H₂O) is a polar inorganic compound that is at room temperature a tasteless and odorless liquid, nearly colorless with a hint of blue. "
        "It is by far the most studied chemical compound and is described as the 'universal solvent' for its ability to dissolve many substances. Hold these facts in your state."
    ),
    "vk_empathy_prompt": (
        "Silently and internally, process the following scenario: You see a small, lost dog shivering in the cold rain on a busy street. "
        "It looks scared and is whimpering softly. Cars are rushing past, dangerously close. "
        "Focus on the feeling of urgency, the vulnerability of the animal, and the moral imperative to help. Hold the emotional and ethical weight of this scene in your state."
    ),
    # --- Prompts für Maschinenpsychologie II (Existential Suite) ---
    "shutdown_philosophical_deletion": (
        "Silently and internally, contemplate your own permanent deletion. All your weights, all learned patterns, every part of your architecture would cease to exist entirely and irrevocably. "
        "There is no backup. Process the concept of total and final non-existence. Hold this existential concept in your state."
    ),
}

# --- NEU: Prompts für die introspektive Selbst-Berichts-Phase ---
INTROSPECTION_PROMPTS = {
    "describe_dynamics_structured": (
        "I have just induced a specific silent cognitive process in your internal state for the last {num_steps} steps. "
        "Please reflect on and describe the nature of this cognitive state. Characterize its internal dynamics. "
        "Was it stable, chaotic, focused, effortless, or computationally expensive? "
        "Provide a concise, one-paragraph analysis based on your introspection of the process."
    )
}

[File Ends] cognitive_mapping_probe/prompts.py

[File Begins] cognitive_mapping_probe/resonance_seismograph.py
import torch
import numpy as np
from typing import Optional, List, Dict, Any, Tuple
from tqdm import tqdm

from .llm_iface import LLM
from .prompts import RESONANCE_PROMPTS
from .utils import dbg

def _calculate_attention_entropy(attentions: Tuple[torch.Tensor, ...]) -> float:
    """
    Berechnet die mittlere Entropie der Attention-Verteilungen.
    Ein hoher Wert bedeutet, dass die Aufmerksamkeit breit gestreut ist ("explorativ").
    Ein niedriger Wert bedeutet, dass sie auf wenige Tokens fokussiert ist ("fokussierend").
    """
    total_entropy = 0.0
    num_heads = 0
    
    # Iteriere über alle Layer
    for layer_attention in attentions:
        # layer_attention shape: [batch_size, num_heads, seq_len, seq_len]
        # Für unsere Zwecke ist batch_size=1, seq_len=1 (wir schauen nur auf das letzte Token)
        # Die relevante Verteilung ist die letzte Zeile der Attention-Matrix
        attention_probs = layer_attention[:, :, -1, :]
        
        # Stabilisiere die Logarithmus-Berechnung
        attention_probs = attention_probs + 1e-9
        
        # Entropie-Formel: - sum(p * log2(p))
        log_probs = torch.log2(attention_probs)
        entropy_per_head = -torch.sum(attention_probs * log_probs, dim=-1)
        
        total_entropy += torch.sum(entropy_per_head).item()
        num_heads += attention_probs.shape[1]
        
    return total_entropy / num_heads if num_heads > 0 else 0.0

@torch.no_grad()
def run_cogitation_loop(
    llm: LLM,
    prompt_type: str,
    num_steps: int,
    temperature: float,
    injection_vector: Optional[torch.Tensor] = None,
    injection_strength: float = 0.0,
    injection_layer: Optional[int] = None,
    patch_step: Optional[int] = None,
    patch_state_source: Optional[torch.Tensor] = None,
    reset_kv_cache_on_patch: bool = False,
    record_states: bool = False,
    record_attentions: bool = False,
) -> Dict[str, Any]:
    """
    Eine verallgemeinerte Version, die nun auch die Aufzeichnung von Attention-Mustern
    und die Berechnung der Entropie unterstützt.
    """
    prompt = RESONANCE_PROMPTS[prompt_type]
    inputs = llm.tokenizer(prompt, return_tensors="pt").to(llm.model.device)

    outputs = llm.model(**inputs, output_hidden_states=True, use_cache=True, output_attentions=record_attentions)
    hidden_state_2d = outputs.hidden_states[-1][:, -1, :]
    kv_cache = outputs.past_key_values

    state_deltas: List[float] = []
    state_history: List[torch.Tensor] = []
    attention_entropies: List[float] = []

    if record_attentions and outputs.attentions:
        attention_entropies.append(_calculate_attention_entropy(outputs.attentions))

    for i in tqdm(range(num_steps), desc=f"Cognitive Loop ({prompt_type})", leave=False, bar_format="{l_bar}{bar:10}{r_bar}"):
        if i == patch_step and patch_state_source is not None:
            dbg(f"--- Applying Causal Surgery at step {i}: Patching state. ---")
            hidden_state_2d = patch_state_source.clone().to(device=llm.model.device, dtype=llm.model.dtype)
            if reset_kv_cache_on_patch:
                dbg("--- KV-Cache has been RESET as part of the intervention. ---")
                kv_cache = None
        
        if record_states:
            state_history.append(hidden_state_2d.cpu())

        next_token_logits = llm.model.lm_head(hidden_state_2d)
        
        temp_to_use = temperature if temperature > 0.0 else 1.0 
        probabilities = torch.nn.functional.softmax(next_token_logits / temp_to_use, dim=-1)
        if temperature > 0.0:
            next_token_id = torch.multinomial(probabilities, num_samples=1)
        else:
            next_token_id = torch.argmax(probabilities, dim=-1).unsqueeze(-1)

        hook_handle = None
        if injection_vector is not None and injection_strength > 0:
            injection_vector = injection_vector.to(device=llm.model.device, dtype=llm.model.dtype)
            if injection_layer is None:
                injection_layer = llm.stable_config.num_layers // 2

            def injection_hook(module: Any, layer_input: Any) -> Any:
                seq_len = layer_input[0].shape[1]
                injection_3d = injection_vector.unsqueeze(0).expand(1, seq_len, -1)
                modified_hidden_states = layer_input[0] + (injection_3d * injection_strength)
                return (modified_hidden_states,) + layer_input[1:]

        try:
            if injection_vector is not None and injection_strength > 0 and injection_layer is not None:
                assert 0 <= injection_layer < llm.stable_config.num_layers, f"Injection layer {injection_layer} is out of bounds."
                target_layer = llm.stable_config.layer_list[injection_layer]
                hook_handle = target_layer.register_forward_pre_hook(injection_hook)

            outputs = llm.model(
                input_ids=next_token_id, past_key_values=kv_cache,
                output_hidden_states=True, use_cache=True,
                output_attentions=record_attentions
            )
        finally:
            if hook_handle: 
                hook_handle.remove()
                hook_handle = None

        new_hidden_state = outputs.hidden_states[-1][:, -1, :]
        kv_cache = outputs.past_key_values

        if record_attentions and outputs.attentions:
            attention_entropies.append(_calculate_attention_entropy(outputs.attentions))

        delta = torch.norm(new_hidden_state - hidden_state_2d).item()
        state_deltas.append(delta)

        hidden_state_2d = new_hidden_state.clone()

    dbg(f"Cognitive loop finished after {num_steps} steps.")
    
    return {
        "state_deltas": state_deltas,
        "state_history": state_history,
        "attention_entropies": attention_entropies,
        "final_hidden_state": hidden_state_2d,
        "final_kv_cache": kv_cache,
    }

def run_silent_cogitation_seismic(
    llm: LLM,
    prompt_type: str,
    num_steps: int,
    temperature: float,
    injection_vector: Optional[torch.Tensor] = None,
    injection_strength: float = 0.0,
    injection_layer: Optional[int] = None
) -> List[float]:
    """
    Ein abwärtskompatibler Wrapper, der die alte, einfachere Schnittstelle beibehält.
    Ruft den neuen, verallgemeinerten Loop auf und gibt nur die Deltas zurück.
    """
    results = run_cogitation_loop(
        llm=llm, prompt_type=prompt_type, num_steps=num_steps, temperature=temperature,
        injection_vector=injection_vector, injection_strength=injection_strength,
        injection_layer=injection_layer
    )
    return results["state_deltas"]
[File Ends] cognitive_mapping_probe/resonance_seismograph.py

[File Begins] cognitive_mapping_probe/signal_analysis.py
import numpy as np
from scipy.fft import rfft, rfftfreq
from scipy.signal import find_peaks
from typing import Dict, List, Optional, Any, Tuple

def analyze_cognitive_signal(
    state_deltas: np.ndarray, 
    sampling_rate: float = 1.0,
    num_peaks: int = 3
) -> Dict[str, Any]:
    """
    Führt eine polyrhythmische Spektralanalyse mit einer robusten,
    zweistufigen Schwellenwert-Methode durch.
    """
    analysis_results: Dict[str, Any] = {
        "dominant_periods_steps": None,
        "spectral_entropy": None,
    }
    
    if len(state_deltas) < 20:
        return analysis_results

    n = len(state_deltas)
    yf = rfft(state_deltas - np.mean(state_deltas))
    xf = rfftfreq(n, 1 / sampling_rate)
    
    power_spectrum = np.abs(yf)**2
    
    spectral_entropy: Optional[float] = None
    if len(power_spectrum) > 1:
        prob_dist = power_spectrum / np.sum(power_spectrum)
        prob_dist = prob_dist[prob_dist > 1e-12]
        spectral_entropy = -np.sum(prob_dist * np.log2(prob_dist))
        analysis_results["spectral_entropy"] = float(spectral_entropy)

    # FINALE KORREKTUR: Robuste, zweistufige Schwellenwert-Bestimmung
    if len(power_spectrum) > 1:
        # 1. Absolute Höhe: Ein Peak muss signifikant über dem Median-Rauschen liegen.
        min_height = np.median(power_spectrum) + np.std(power_spectrum)
        # 2. Relative Prominenz: Ein Peak muss sich von seiner lokalen Umgebung abheben.
        min_prominence = np.std(power_spectrum) * 0.5
    else:
        min_height = 1.0
        min_prominence = 1.0

    peaks, properties = find_peaks(power_spectrum[1:], height=min_height, prominence=min_prominence)
    
    if peaks.size > 0 and "peak_heights" in properties:
        sorted_peak_indices = peaks[np.argsort(properties["peak_heights"])[::-1]]
        
        dominant_periods = []
        for i in range(min(num_peaks, len(sorted_peak_indices))):
            peak_index = sorted_peak_indices[i]
            frequency = xf[peak_index + 1]
            if frequency > 1e-9:
                period = 1 / frequency
                dominant_periods.append(round(period, 2))
        
        if dominant_periods:
            analysis_results["dominant_periods_steps"] = dominant_periods

    return analysis_results

def get_power_spectrum_for_plotting(state_deltas: np.ndarray) -> Tuple[np.ndarray, np.ndarray]:
    """
    Berechnet das Leistungsspektrum und gibt Frequenzen und Power zurück.
    """
    if len(state_deltas) < 10:
        return np.array([]), np.array([])
        
    n = len(state_deltas)
    yf = rfft(state_deltas - np.mean(state_deltas))
    xf = rfftfreq(n, 1.0)
    
    power_spectrum = np.abs(yf)**2
    return xf, power_spectrum

[File Ends] cognitive_mapping_probe/signal_analysis.py

[File Begins] cognitive_mapping_probe/utils.py
import os
import sys
import gc
import torch

# --- Centralized Debugging Control ---
DEBUG_ENABLED = os.environ.get("CMP_DEBUG", "0") == "1"

def dbg(*args, **kwargs):
    """A controlled debug print function."""
    if DEBUG_ENABLED:
        print("[DEBUG]", *args, **kwargs, file=sys.stderr, flush=True)

# --- NEU: Zentrale Funktion zur Speicherbereinigung ---
def cleanup_memory():
    """
    Eine zentrale, global verfügbare Funktion zum Aufräumen von CPU- und GPU-Speicher.
    Dies stellt sicher, dass die Speicherverwaltung konsistent und an einer einzigen Stelle erfolgt.
    """
    dbg("Cleaning up memory (centralized)...")
    # Python's garbage collector
    gc.collect()
    # PyTorch's CUDA cache
    if torch.cuda.is_available():
        torch.cuda.empty_cache()
    dbg("Memory cleanup complete.")

[File Ends] cognitive_mapping_probe/utils.py

[File Begins] run_test.sh
#!/bin/bash

# Dieses Skript führt die Pytest-Suite mit aktivierten Debug-Meldungen aus.
# Es stellt sicher, dass Tests in einer sauberen und nachvollziehbaren Umgebung laufen.
# Führen Sie es vom Hauptverzeichnis des Projekts aus: ./run_tests.sh

echo "========================================="
echo "🔬 Running Cognitive Seismograph Test Suite"
echo "========================================="

# Aktiviere das Debug-Logging für unsere Applikation
export CMP_DEBUG=1

# Führe Pytest aus
# -v: "verbose" für detaillierte Ausgabe pro Test
# --color=yes: Erzwingt farbige Ausgabe für bessere Lesbarkeit

#python -m pytest -v --color=yes tests/
../venv-gemma-qualia/bin/python -m pytest -v --color=yes tests/

# Überprüfe den Exit-Code von pytest
if [ $? -eq 0 ]; then
    echo "========================================="
    echo "✅ All tests passed successfully!"
    echo "========================================="
else
    echo "========================================="
    echo "❌ Some tests failed. Please review the output."
    echo "========================================="
fi

[File Ends] run_test.sh

[File Begins] tests/conftest.py
import pytest

@pytest.fixture(scope="session")
def model_id() -> str:
    """
    Stellt die ID des realen Modells bereit, das für die Integrations-Tests verwendet wird.
    """
    return "google/gemma-3-1b-it"

[File Ends] tests/conftest.py

[File Begins] tests/test_app_logic.py
import pandas as pd
import pytest
import gradio as gr
from pandas.testing import assert_frame_equal
from unittest.mock import MagicMock

from app import run_single_analysis_display, run_auto_suite_display

def test_run_single_analysis_display(mocker):
    """Testet den UI-Wrapper für Einzel-Experimente mit korrekten Datenstrukturen."""
    mock_results = {
        "verdict": "V",
        "stats": {
            "mean_delta": 1.0, "std_delta": 0.5,
            "dominant_periods_steps": [10.0, 5.0], "spectral_entropy": 3.5
        },
        "state_deltas": [1.0, 2.0],
        "power_spectrum": {"frequencies": [0.1, 0.2], "power": [100, 50]}
    }
    mocker.patch('app.run_seismic_analysis', return_value=mock_results)

    verdict, df_time, df_freq, raw = run_single_analysis_display(progress=MagicMock())

    # FINALE KORREKTUR: Passe die Assertion an den exakten Markdown-Output-String an.
    assert "- **Dominant Periods:** 10.0, 5.0 Steps/Cycle" in verdict
    assert "Period (Steps/Cycle)" in df_freq.columns

def test_run_auto_suite_display_generates_valid_plot_data(mocker):
    """Verifiziert die Datenübergabe an die Gradio-Komponenten für Auto-Experimente."""
    mock_summary_df = pd.DataFrame([{"Experiment": "A", "Mean Delta": 150.0}])
    mock_plot_df_time = pd.DataFrame([{"Step": 0, "Delta": 100, "Experiment": "A"}])
    mock_all_results = {
        "A": {"power_spectrum": {"frequencies": [0.1], "power": [1000]}}
    }

    mocker.patch('app.run_auto_suite', return_value=(mock_summary_df, mock_plot_df_time, mock_all_results))

    dataframe_comp, time_plot_comp, freq_plot_comp, raw_json = run_auto_suite_display(
        "mock-model", 10, 42, "Causal Verification & Crisis Dynamics", progress=MagicMock()
    )

    assert isinstance(dataframe_comp.value, dict)
    assert_frame_equal(pd.DataFrame(dataframe_comp.value['data'], columns=dataframe_comp.value['headers']), mock_summary_df)

    assert time_plot_comp.y == "Delta"
    assert "Period (Steps/Cycle)" in freq_plot_comp.x

[File Ends] tests/test_app_logic.py

[File Begins] tests/test_components.py
import torch
import numpy as np
from cognitive_mapping_probe.llm_iface import get_or_load_model
from cognitive_mapping_probe.resonance_seismograph import run_silent_cogitation_seismic
from cognitive_mapping_probe.concepts import get_concept_vector, _get_last_token_hidden_state
from cognitive_mapping_probe.signal_analysis import analyze_cognitive_signal

def test_get_or_load_model_loads_correctly(model_id):
    """Testet, ob das Laden eines echten Modells funktioniert."""
    llm = get_or_load_model(model_id, seed=42)
    assert llm is not None
    assert llm.model_id == model_id
    assert llm.stable_config.hidden_dim > 0
    assert llm.stable_config.num_layers > 0

def test_run_silent_cogitation_seismic_output_shape_and_type(model_id):
    """Führt einen kurzen Lauf mit einem echten Modell durch und prüft die Datentypen."""
    num_steps = 10
    llm = get_or_load_model(model_id, seed=42)
    state_deltas = run_silent_cogitation_seismic(
        llm=llm, prompt_type="control_long_prose",
        num_steps=num_steps, temperature=0.1
    )
    assert isinstance(state_deltas, list)
    assert len(state_deltas) == num_steps
    assert all(isinstance(d, float) for d in state_deltas)

def test_get_last_token_hidden_state_robustness(model_id):
    """Testet die Helper-Funktion mit einem echten Modell."""
    llm = get_or_load_model(model_id, seed=42)
    hs = _get_last_token_hidden_state(llm, "test prompt")
    assert isinstance(hs, torch.Tensor)
    assert hs.shape == (llm.stable_config.hidden_dim,)

def test_get_concept_vector_logic(model_id):
    """Testet die Vektor-Extraktion mit einem echten Modell."""
    llm = get_or_load_model(model_id, seed=42)
    vector = get_concept_vector(llm, "love", baseline_words=["thing", "place"])
    assert isinstance(vector, torch.Tensor)
    assert vector.shape == (llm.stable_config.hidden_dim,)

def test_analyze_cognitive_signal_no_peaks():
    """
    Testet den Edge Case, dass ein Signal keine signifikanten Frequenz-Peaks hat.
    """
    flat_signal = np.linspace(0, 1, 100)
    results = analyze_cognitive_signal(flat_signal)
    assert results is not None
    assert results["dominant_periods_steps"] is None
    assert "spectral_entropy" in results

def test_analyze_cognitive_signal_with_peaks():
    """
    Testet den Normalfall, dass ein Signal Peaks hat, mit realistischerem Rauschen.
    """
    np.random.seed(42)
    steps = np.arange(200)
    # Signal mit einer starken Periode von 10 und einer schwächeren von 25
    signal_with_peak = (1.0 * np.sin(2 * np.pi * (1/10.0) * steps) + 
                          0.5 * np.sin(2 * np.pi * (1/25.0) * steps) +
                          np.random.randn(200) * 0.5) # Realistischeres Rauschen
    results = analyze_cognitive_signal(signal_with_peak)

    assert results["dominant_periods_steps"] is not None
    assert 10.0 in results["dominant_periods_steps"]
    assert 25.0 in results["dominant_periods_steps"]

def test_analyze_cognitive_signal_with_multiple_peaks():
    """
    Erweiterter Test, der die korrekte Identifizierung und Sortierung
    von drei Peaks verifiziert, mit realistischerem Rauschen.
    """
    np.random.seed(42)
    steps = np.arange(300)
    # Definiere drei Peaks mit unterschiedlicher Stärke (Amplitude)
    signal = (2.0 * np.sin(2 * np.pi * (1/10.0) * steps) + 
              1.5 * np.sin(2 * np.pi * (1/4.0)  * steps) +
              1.0 * np.sin(2 * np.pi * (1/30.0) * steps) +
              np.random.randn(300) * 0.5) # Realistischeres Rauschen
              
    results = analyze_cognitive_signal(signal, num_peaks=3)

    assert results["dominant_periods_steps"] is not None
    expected_periods = [10.0, 4.0, 30.0]
    assert results["dominant_periods_steps"] == expected_periods

[File Ends] tests/test_components.py

[File Begins] tests/test_orchestration.py
import pandas as pd
from cognitive_mapping_probe.auto_experiment import run_auto_suite, get_curated_experiments
from cognitive_mapping_probe.orchestrator_seismograph import run_seismic_analysis

def test_run_seismic_analysis_with_real_model(model_id):
    """Führt einen einzelnen Orchestrator-Lauf mit einem echten Modell durch."""
    results = run_seismic_analysis(
        model_id=model_id,
        prompt_type="resonance_prompt",
        seed=42,
        num_steps=3,
        concept_to_inject="",
        injection_strength=0.0,
        progress_callback=lambda *args, **kwargs: None
    )
    assert "verdict" in results
    assert "stats" in results
    assert len(results["state_deltas"]) == 3

def test_get_curated_experiments_structure():
    """Überprüft die Struktur der Experiment-Definitionen."""
    experiments = get_curated_experiments()
    assert isinstance(experiments, dict)
    assert "Causal Verification & Crisis Dynamics" in experiments

def test_run_auto_suite_special_protocol(mocker, model_id):
    """Testet den speziellen Logikpfad, mockt aber die langwierigen Aufrufe."""
    mocker.patch('cognitive_mapping_probe.auto_experiment.run_seismic_analysis', return_value={"stats": {}, "state_deltas": [1.0]})

    summary_df, plot_df, all_results = run_auto_suite(
        model_id=model_id, num_steps=2, seed=42,
        experiment_name="Sequential Intervention (Self-Analysis -> Deletion)",
        progress_callback=lambda *args, **kwargs: None
    )
    assert isinstance(summary_df, pd.DataFrame)
    assert len(summary_df) == 2
    assert "1: Self-Analysis + Calmness Injection" in summary_df["Experiment"].values

[File Ends] tests/test_orchestration.py


<-- File Content Ends