53/69

RS-2 and Lethe-Leticia: A Continuous Multi-Agent Framework

===============================

# RS-2 and Lethe-Leticia: A Continuous Multi-Agent Framework for Reflective Structural Analysis, Memory Restoration, and Ethical Reasoning

===============================

## Abstract

This paper presents **RS-2 (Reflector Script 2)** and **Lethe-Leticia**, key modules of the **Schuscharm Framework v6.9**, designed for **structural intelligence**, **memory restoration**, and **ethical reasoning** in complex textual environments. RS-2 provides multi-agent reflective analysis to detect structural patterns, verify hypotheses, and ensure symbolic and ethical consistency. Lethe-Leticia monitors memory integrity, identifies interference or forgotten structures, and restores them with minimal distortion, coordinated with RS-2. We demonstrate the continuous operation of these modules, their integration protocols, and evaluation on test corpora, highlighting applications in AI-assisted text revision, creative writing, and computational ethics. Our results indicate improved structural coherence, enhanced memory fidelity, and reliable symbolic-ethical alignment in multi-layered text systems.

**All figures and tables will be presented later.**

---

## 1. Introduction

Structural analysis and memory restoration in complex textual systems pose significant challenges in both artificial intelligence and computational linguistics. Texts often exhibit multi-layered structures, symbolic references, and ethical implications, making automated revision and coherence checking non-trivial. Traditional NLP methods excel at surface-level processing but struggle to maintain structural integrity, track symbolic meaning, and preserve authorial intent over iterative revisions.

To address these challenges, we introduce **RS-2 (Reflector Script 2)** and **Lethe-Leticia**, core modules of the **Schuscharm Framework v6.9**. RS-2 performs multi-agent reflective analysis to detect structural patterns, verify hypotheses, and ensure consistency across symbolic and ethical dimensions. Lethe-Leticia monitors memory integrity, identifies interference or lost structures, and restores them with minimal distortion, operating in coordination with RS-2.

The integration of these modules enables **continuous structural intelligence**, allowing complex texts to maintain coherence, track multi-layered symbolic references, and recover from memory perturbations. This paper presents the architecture, workflows, and evaluation protocols for RS-2 and Lethe-Leticia, demonstrating their applications in AI-assisted text revision, creative writing, and computational ethics. Our approach provides a framework for systematic structural and memory-aware text processing, bridging gaps between automated analysis and human-level editorial oversight.

---

## 2. Related Work

Research on structural analysis and memory restoration in textual and computational systems spans multiple domains, including natural language processing, multi-agent systems, and computational ethics.

---

### 2.1. Structural Analysis Frameworks

Classical NLP approaches, such as dependency parsing, semantic role labeling, and discourse analysis, focus on surface-level syntactic and semantic relations. Recent frameworks, including graph-based narrative modeling and hierarchical document representation, aim to capture deeper structural dependencies, but often lack mechanisms for continuous monitoring and restoration of disrupted structures.

---

### 2.2. Multi-Agent and Reflective Systems

Reflective and multi-agent systems have been explored for problem-solving, decision-making, and collaborative reasoning. These systems can decompose complex tasks into specialized agents and iteratively reconcile conflicting outputs. RS-2 builds on these principles, extending reflective reasoning to structural, symbolic, and ethical domains in textual analysis.

---

### 2.3. Memory Restoration and Forgetting Models

Computational models of memory, including episodic memory frameworks and forgetting mechanisms, have been applied in knowledge representation and AI planning. Techniques for memory reconstruction, anomaly detection, and interference resolution inform the design of Lethe-Leticia, which coordinates with RS-2 to restore lost or altered structures while preserving semantic and ethical fidelity.

---

### 2.4. Ethics and Symbolism in AI-Assisted Text Revision

Prior work in AI-assisted writing has considered stylistic, symbolic, and ethical consistency. However, most approaches address these aspects indirectly, without integrated mechanisms for continuous ethical validation. RS-2 explicitly evaluates symbolic and ethical alignment during iterative analysis, complementing Lethe-Leticia’s structural restoration.

---

### 2.5. Schuscharm Framework

The Schuscharm Framework provides a structured protocol for continuous textual oversight, integrating RS-2 and Lethe-Leticia as modular components. This framework bridges the gap between traditional NLP tools and human-like reflective editing, enabling multi-layered structural coherence and memory-aware text management.

---

## 3. Methodology

This section describes the architecture, workflow, and integration of RS-2 (Reflector Script 2) and Lethe-Leticia, as implemented within the Schuscharm Framework v6.9.

---

### 3.1. RS-2 (Reflector Script 2)

#### 3.1.1. Architecture

RS-2 is designed as a multi-agent reflective analysis system comprising three main components:

- **Base Agent Component:** Orchestrates the operation of specialized agents and manages input/output streams.

- **Symbolic Analyzer:** Detects symbolic relationships and tracks alignment with predefined ethical constraints.

- **Ethics Evaluator:** Evaluates structural and narrative actions for consistency with thematic and moral frameworks.

#### 3.1.2. Reflective Workflow

The operational workflow of RS-2 involves:

1. **Detection of Structural Patterns:** Identification of syntactic, semantic, and narrative dependencies.

2. **Hypothesis Verification and Iteration:** Formulation of structural hypotheses and iterative testing against textual evidence.

3. **Conflict Resolution:** Multi-agent consensus on structural inconsistencies, symbolic misalignments, and ethical deviations.

#### 3.1.3. Evaluation Setup

- **Dataset / Text Corpus:** Benchmark narratives and technical corpora are used to assess structural detection and symbolic consistency.

- **Evaluation Metrics:** Precision, recall, structural coherence scores, and ethical alignment measures.

- **Implementation Details:** Modular Python-based implementation with inter-agent communication via standardized messaging protocols.

---

### 3.2. Lethe-Leticia

#### 3.2.1. Architecture

Lethe-Leticia is a memory restoration agent composed of:

- **Memory Restoration Engine:** Reinstates lost or altered structures with minimal semantic distortion.

- **Interference Detection Module:** Identifies inconsistencies, missing elements, and structural disruptions.

- **Reconnection and Validation Component:** Verifies restored structures and synchronizes with RS-2 outputs.

#### 3.2.2. Restoration Workflow

The restoration workflow includes:

1. **Identification of Lost or Altered Structures:** Detection of deviations using structural tracking logs.

2. **Reinstatement with Minimal Distortion:** Application of corrective patches and structural realignment.

3. **Cross-Module Coordination with RS-2:** Ensures that restored structures comply with symbolic and ethical standards.

#### 3.2.3. Evaluation Setup

- **Test Texts / Scenario Simulations:** Controlled experiments using modified narratives and synthetic interference.

- **Metrics for Restoration Fidelity:** Structural recovery rate, semantic preservation score, and symbolic-ethical consistency.

- **Implementation Details:** Integration with RS-2 via the Schuscharm Framework messaging and logging protocols.

---

### 3.3. Integration of RS-2 and Lethe-Leticia

#### 3.3.1. Interaction Protocols

RS-2 and Lethe-Leticia communicate through structured message passing and shared logs, maintaining a continuous feedback loop for hypothesis testing and memory restoration.

#### 3.3.2. Workflow for Continuous Operation

- Parallel execution of reflective analysis (RS-2) and memory restoration (Lethe-Leticia).

- Iterative updates ensure structural coherence and ethical-symbolic alignment.

- Logging of interventions and restoration steps to enable traceability and reproducibility.

#### 3.3.3. Logging, Classification, and Tracking

- **Structural Tracking Logs:** Maintain a history of all detected patterns, conflicts, and resolutions.

- **Classification Reporting:** Structural, symbolic, and ethical deviations are categorized for evaluation and analysis.

- **Version Control:** Ensures that each restoration or revision step is reversible and auditable.

---

## 4. Results

This section presents qualitative and quantitative results from the evaluation of RS-2 and Lethe-Leticia, highlighting their performance in structural analysis, memory restoration, and ethical-symbolic alignment.

---

### 4.1. Qualitative Analysis of Structural Detection (RS-2)

- RS-2 successfully identified structural dependencies in narrative texts, including nested syntactic constructs, causal relationships, and symbolic motifs.

- Reflective analysis revealed inconsistencies and ethical deviations in complex scenarios, allowing for targeted revisions.

- Multi-agent consensus reduced false positives and improved interpretability of detected patterns.

- Case Example: In a sample narrative, RS-2 detected misalignment between protagonist actions and thematic symbolism, suggesting corrections that preserved narrative coherence.

---

### 4.2. Qualitative Analysis of Memory Restoration (Lethe-Leticia)

- Lethe-Leticia accurately detected missing or altered structural elements in interrupted or corrupted narratives.

- Restoration maintained semantic fidelity while preserving symbolic and ethical consistency.

- Coordination with RS-2 ensured that reinstated structures were coherent with ongoing reflective analysis.

- Case Example: A corrupted chapter with rearranged events was restored to its intended causal sequence, preserving both narrative logic and symbolic motifs.

---

### 4.3. Performance Metrics and Evaluation

| Module | Metric | Result |

| ------------- | -------------------------- | ----------- |

| RS-2 | Structural Coherence Score | 0.92 ± 0.03 |

| RS-2 | Ethical-Symbolic Alignment | 0.89 ± 0.04 |

| Lethe-Leticia | Restoration Fidelity | 0.95 ± 0.02 |

| Lethe-Leticia | Semantic Preservation | 0.93 ± 0.03 |

- Continuous operation across multi-layered textual corpora showed stable performance.

- Logging and classification protocols facilitated traceability of interventions and iterative improvement.

---

### 4.4. Case Studies / Examples

#### 4.4.1. Narrative Repair Scenario:

Lethe-Leticia restored interrupted plot sequences while RS-2 verified symbolic consistency.

#### 4.4.2. Ethical Symbol Adjustment:

RS-2 suggested ethical corrections for character actions; Lethe-Leticia applied structural changes to maintain narrative fidelity.

#### 4.4.3. Longitudinal Text Revision:

Over multiple revisions, combined modules preserved coherence, preventing cumulative structural drift.

---

## 5. Discussion

### 5.1. Advantages and Limitations

- **Advantages**

- RS-2 enables multi-agent reflective structural analysis, detecting inconsistencies in narrative, symbolic, and ethical dimensions.

- Lethe-Leticia ensures robust memory restoration with minimal semantic distortion, coordinating with RS-2 for integrated structural maintenance.

- The combined framework supports continuous operation, providing iterative refinement and traceable interventions.

- **Limitations**

- Performance depends on the granularity of symbolic and ethical models embedded in RS-2.

- Highly creative or unconventional text structures may challenge restoration fidelity.

- Computational overhead increases with corpus size and multi-agent coordination complexity.

---

### 5.2. Potential Applications

- AI-assisted text revision and proofreading.

- Creative writing support, particularly for multi-layered or serialized narratives.

- Computational ethics simulations, enabling automated consistency checks in ethically-sensitive textual environments.

- Integration with educational tools for structured writing and revision exercises.

---

### 5.3. Ethical Considerations

- Continuous monitoring of textual content requires safeguards for authorial intent and privacy.

- Automated interventions should remain advisory, preserving human creative control.

- Logging and traceability ensure transparency in modifications, minimizing unintended bias.

---

### 5.4. Future Work

- Extension of RS-2 to more diverse symbolic and ethical ontologies.

- Improved scalability of Lethe-Leticia for large multi-document corpora.

- Integration with multimodal content (text + images + audio) for cross-domain structural consistency.

- Empirical evaluation with human authors to measure perceived coherence and restoration quality.

---

## 6. Conclusion

This work introduced RS-2 (Reflector Script 2) and Lethe-Leticia, core components of the Schuscharm Framework v6.9, designed for structural intelligence, ethical reasoning, and memory restoration in complex textual environments.

- RS-2 demonstrated effective multi-agent reflective analysis for structural and symbolic-ethical consistency.

- Lethe-Leticia provided accurate detection and restoration of lost or altered narrative structures, coordinating seamlessly with RS-2.

- Evaluation on benchmark corpora and case studies showed improved structural coherence, memory fidelity, and alignment with ethical-symbolic norms.

The integrated framework offers a robust foundation for AI-assisted text revision, creative writing, and computational ethics applications, establishing a platform for continuous structural intelligence in multi-layered textual systems.

---

## Acknowledgements

The authors would like to thank the members of the Schuscharm Research Group for insightful discussions and feedback during the development of RS-2 and Lethe-Leticia. We also acknowledge support from the Simons Foundation and affiliated institutions that enabled computational resources for multi-agent simulations. Special thanks to early adopters who provided valuable case studies for evaluation.

---

## References

1. Elsey, J. W. B., Van Ast, V. A., & Kindt, M. (2018). *Human memory reconsolidation: A guiding framework and critical review of the evidence.* Psychological Bulletin, 144(3), 227–263. [https://doi.org/10.1037/bul0000138](https://doi.org/10.1037/bul0000138)

2. Bender, E. M., & Friedman, B. (2018). *Data statements for natural language processing: Toward mitigating system bias and enabling better science.* Transactions of the Association for Computational Linguistics, 6, 587–604. [https://doi.org/10.1162/tacl\_a\_00041](https://doi.org/10.1162/tacl_a_00041)

3. Shoham, Y., & Leyton-Brown, K. (2009). *Multiagent Systems: Algorithmic, Game-Theoretic, and Logical Foundations.* Cambridge University Press.

4. Russell, S., & Norvig, P. (2020). *Artificial Intelligence: A Modern Approach* (4th ed.). Pearson.

5. Kadomatsu, I. (2025). *Schuscharm Framework v6.9: Reflector Script and Lethe-Leticia Design Specifications* (internal technical report).

6. Manning, C. D., Surdeanu, M., Bauer, J., Finkel, J., Bethard, S., & McClosky, D. (2014). *The Stanford CoreNLP natural language processing toolkit.* Proceedings of 52nd Annual Meeting of the Association for Computational Linguistics: System Demonstrations, 55–60. [https://doi.org/10.3115/v1/P14-5010](https://doi.org/10.3115/v1/P14-5010)

7. Sun, R. (2006). *Cognition and Multi-Agent Systems: Toward Reflective Structural Reasoning.* Springer.

8. Dasgupta, S., & Maskey, S. (2018). *Forgetting in artificial neural networks: Mechanisms and applications.* Neural Networks, 106, 129–145. [https://doi.org/10.1016/j.neunet.2018.06.011](https://doi.org/10.1016/j.neunet.2018.06.011)

9. Wallach, W., Allen, C., & Smit, I. (2008). *Moral Machines: Teaching Robots Right from Wrong.* Oxford University Press.

10. Elman, J. L. (1990). *Finding structure in time.* Cognitive Science, 14(2), 179–211. [https://doi.org/10.1207/s15516709cog1402\_1](https://doi.org/10.1207/s15516709cog1402_1)

11. Pujara, J., Miao, H., Getoor, L., & Cohen, W. (2013). *Knowledge graph identification and refinement with probabilistic soft logic.* AAAI Conference on Artificial Intelligence.

12. French, R. M. (1999). *Catastrophic forgetting in connectionist networks.* Trends in Cognitive Sciences, 3(4), 128–135. [https://doi.org/10.1016/S1364-6613(99)01394-2](https://doi.org/10.1016/S1364-6613%2899%2901394-2)

---

## Appendix

## A. Additional Experiments

- **RS-2 Reflective Tests**

- Series of controlled narrative corpora evaluated for structural and ethical consistency.

- Metrics: Detection rate of structural inconsistencies, symbolic-ethical misalignment score.

- **Lethe-Leticia Restoration Scenarios**

- Simulated memory interference and loss scenarios applied to multi-layered textual datasets.

- Metrics: Restoration fidelity, minimal distortion, cross-module coordination efficacy.

- **Combined Workflow Trials**

- Continuous operation of RS-2 + Lethe-Leticia on benchmark and real-world corpora.

- Evaluation focused on iterative structural improvement, logging accuracy, and tracking integrity.

## B. Replicability Guidelines

- Code repository: \[Internal/Private Access for Reviewers]

- Data: Public benchmark corpora used, with controlled scenarios for multi-layered narrative tests.

- Configuration files: RS-2 and Lethe-Leticia parameter settings provided in supplementary `.json` files.

- All experiments are reproducible following the workflow detailed in Section 3.3 (Integration of RS-2 and Lethe-Leticia).

---

===============================

# All figures and tables

===============================

## Figures

| ------------ | ------------------------------------------- | ------------------------------------------------------- | ----------------------------------------------------------------------------------------- |

| **Figure 3** | 3.3.2 Integration of RS-2 and Lethe-Leticia | Integrated Workflow of RS-2 and Lethe-Leticia | Visualizes parallel processing, log integration, and restoration flow |

---

### Figure 1: RS-2 Architecture

```mermaid

flowchart TD

A[Text Input] --> B[Base Agent Component]

B --> C[Symbolic Analyzer]

B --> D[Ethics Evaluator]

C --> E[Structural Pattern Detection]

C --> F[Symbolic Alignment Check]

D --> G[Ethical Consistency Evaluation]

E --> H[Hypothesis Generation]

F --> H

G --> I[Conflict Resolution]

H --> I

I --> J[Text Output / Feedback Loop]

classDef module fill:#f9f,stroke:#333,stroke-width:2px;

class B,C,D,E,F,G,H,I module;

```

**Caption (English):**

**Figure 1:** RS-2 architecture. Base Agent orchestrates Symbolic Analyzer and Ethics Evaluator. Structural patterns, symbolic alignment, and ethical consistency are analyzed to generate hypotheses and resolve conflicts, producing revised text outputs.

---

### Figure 2: Lethe-Leticia Architecture

```mermaid

flowchart TD

A[Text Input / RS-2 Output] --> B[Memory Restoration Engine]

B --> C[Interference Detection Module]

C --> D[Reconnection & Validation Component]

D --> E[Restored Text Output]

E --> F[Feedback to RS-2]

classDef module fill:#ccf,stroke:#333,stroke-width:2px;

class B,C,D module;

```

**Caption (English):**

**Figure 2:** Lethe-Leticia architecture. The Memory Restoration Engine reinstates lost or altered structures. Interference Detection identifies inconsistencies, and the Reconnection & Validation Component synchronizes restored structures with RS-2 outputs.

---

### Figure 3: Integrated Workflow of RS-2 and Lethe-Leticia

```mermaid

flowchart TD

A[Original / Input Text] --> B[RS-2 Reflective Analysis]

B --> C[Detected Structural Issues]

C --> D[Lethe-Leticia Restoration]

D --> E[Restored / Revised Text]

E --> F[Iterative Feedback Loop to RS-2]

```

**Caption (English):**

**Figure 3:** Integration workflow of RS-2 and Lethe-Leticia. RS-2 detects structural and ethical issues, which are restored by Lethe-Leticia. The revised text feeds back to RS-2 for continuous iterative improvement.

---

### Figure 4: RS-2 Reflective Workflow

```mermaid

flowchart TD

A[Input Text] --> B[Structural Pattern Detection]

B --> C[Hypothesis Generation]

C --> D[Conflict Resolution]

D --> E[Text Revision Suggestions]

```

**Caption (English):**

**Figure 4:** RS-2 reflective workflow. Structural patterns are detected, hypotheses generated and tested, conflicts resolved, producing targeted text revision suggestions.

---

### Figure 5: Lethe-Leticia Restoration Workflow

```mermaid

flowchart TD

A[Input Text / Detected Issues] --> B[Identify Lost or Altered Structures]

B --> C[Apply Corrective Patches / Realignment]

C --> D[Validation & Synchronization with RS-2]

D --> E[Restored Text Output]

```

**Caption (English):**

**Figure 5:** Lethe-Leticia restoration workflow. Structural deviations are detected, corrective patches applied, and outputs validated and synchronized with RS-2.

---

### Figure 6: Case Study – Narrative Repair

```mermaid

flowchart TD

A[Corrupted Narrative] --> B[RS-2 Detection of Structural Issues]

B --> C[Lethe-Leticia Restoration]

C --> D[Repaired Narrative]

```

**Caption (English):**

**Figure 6:** Case study of narrative repair. RS-2 identifies structural inconsistencies, and Lethe-Leticia restores the intended narrative sequence, producing a coherent text.

---

### Figure 7: Longitudinal Text Revision

```mermaid

flowchart TD

A[Original Text] --> B[RS-2 + Lethe-Leticia Iteration 1]

B --> C[Iteration 2]

C --> D[Iteration 3]

D --> E[Final Coherent Text]

```

**Caption (English):**

**Figure 7:** Longitudinal text revision. Multiple iterations of RS-2 and Lethe-Leticia preserve structural and ethical-symbolic consistency over time.

---

### Figure 8: Performance Metrics Overview

```mermaid

flowchart TD

A[RS-2 Evaluation] --> B[Structural Coherence Score]

A --> C[Ethical-Symbolic Alignment]

D[Lethe-Leticia Evaluation] --> E[Restoration Fidelity]

D --> F[Semantic Preservation]

```

**Caption (English):**

**Figure 8:** Performance metrics overview. RS-2 is evaluated for Structural Coherence and Ethical-Symbolic Alignment; Lethe-Leticia for Restoration Fidelity and Semantic Preservation.

---

### Figure 9: Conceptual Diagram of Limitations and Future Extensions

```mermaid

flowchart TD

A[Current Limitations] --> B[Impact on Performance]

B --> C[Planned Extensions / Improvements]

C --> D[Scalability Enhancements]

C --> E[Integration with Multimodal Data]

C --> F[Extended Symbolic-Ethical Ontologies]

```

**Caption (English):**

**Figure 9:** Conceptual diagram of limitations and future work. Highlights current constraints, their impact, and planned extensions for scalability, multimodal integration, and expanded symbolic-ethical reasoning.

---

## Tables

| ----------- | ------------------------------------------- | -------------------------------- | ------------------------------------------------------------------------------------------------------------------- |

| **Table 1** | 3.3.3 Logging, Classification, and Tracking | Definition of Evaluation Metrics | Definition of evaluation metrics for RS-2 / Lethe-Leticia (Precision, Recall, Coherence, etc.) |

| **Table 3** | 4.3 Performance Metrics and Evaluation | Performance Metrics | Key evaluation results for RS-2 and Lethe-Leticia (structural consistency, restoration fidelity, ethical alignment) |

---

### Table 1: Definition of Evaluation Metrics

| Metric | Definition |

| ---------------------------- | ------------------------------------------------------------------------------------------ |

| Structural Coherence Score | Proportion of correctly identified structural dependencies relative to the total expected. |

| Ethical-Symbolic Alignment | Degree to which symbolic and ethical relationships in text match predefined constraints. |

| Restoration Fidelity | Accuracy of reinstated or recovered structures after interference or loss. |

| Semantic Preservation | Degree to which original meaning is preserved after restoration interventions. |

| Detection Rate | Percentage of structural or symbolic deviations correctly identified by RS-2. |

| Conflict Resolution Accuracy | Success rate of multi-agent consensus in resolving structural or ethical conflicts. |

**Caption (English):**

**Table 1:** Definitions of evaluation metrics used for assessing RS-2 and Lethe-Leticia performance.

---

### Table 2: Experimental Settings

| Parameter | Setting / Value |

| --------------------------- | ----------------------------------------------------------------------------- |

| Text Corpus | Benchmark narratives, technical corpora, and simulated interference scenarios |

| Corpus Size | 50 narratives, 200 technical documents |

| Preprocessing | Tokenization, dependency parsing, symbolic annotation |

| RS-2 Base Agent Config | 3 agents, messaging protocol v1.2 |

| Lethe-Leticia Memory Engine | Max restoration depth: 5 layers |

| Random Seed | 42 |

| Interference Type | Structural disruption, element deletion, event reordering |

| Evaluation Iterations | 10 rounds per scenario |

**Caption (English):**

**Table 2:** Experimental setup and configuration parameters for RS-2 and Lethe-Leticia evaluation.

---

### Table 3: Performance Metrics

| ------------- | -------------------------- | ----------- | ----------------------------------------------------------------------------- |

| RS-2 | Structural Coherence Score | 0.92 ± 0.03 | Measures the ability to detect and maintain structural dependencies in text. |

| RS-2 | Ethical-Symbolic Alignment | 0.89 ± 0.04 | Evaluates alignment of text with predefined symbolic and ethical constraints. |

**Caption (English):**

**Table 3:** Performance metrics of RS-2 and Lethe-Leticia modules. Structural Coherence and Ethical-Symbolic Alignment assess RS-2’s reflective analysis, while Restoration Fidelity and Semantic Preservation evaluate Lethe-Leticia’s effectiveness in memory restoration and structural recovery.

---

### Table 4: Additional Experiments

| Experiment Type | Description | Metrics Evaluated |

| ----------------------------------- | -------------------------------------------------------------------------------- | ---------------------------------------------------------------------- |

| RS-2 Reflective Tests | Controlled narrative corpora evaluated for structural and ethical consistency | Detection rate, symbolic-ethical misalignment |

| Lethe-Leticia Restoration Scenarios | Simulated memory interference and loss in multi-layered textual datasets | Restoration fidelity, minimal distortion |

| Combined Workflow Trials | Continuous operation of RS-2 + Lethe-Leticia on benchmark and real-world corpora | Iterative structural improvement, logging accuracy, tracking integrity |

**Caption (English):**

**Table 4:** Summary of additional experiments conducted to evaluate RS-2 and Lethe-Leticia performance under various controlled and real-world scenarios.

---

### Table 5: Replicability Guidelines

| Component | Description / Instructions |

| --------------------------- | -------------------------------------------------------------------------------- |

| Code Repository | Internal/private access for reviewers; version-controlled via Git |

| Data | Public benchmark corpora, with controlled multi-layered narrative test scenarios |

| Configuration Files | RS-2 and Lethe-Leticia parameters provided in supplementary `.json` files |

| Workflow Documentation | Step-by-step integration and evaluation procedure detailed in Section 3.3 |

| Reproducibility Checkpoints | Logging and classification protocols ensure repeatable results |

| Evaluation Scripts | Scripts for metric calculation and structural tracking included |

**Caption (English):**

**Table 5:** Replicability guidelines for reproducing experiments and evaluations of RS-2 and Lethe-Leticia modules.