Sitemap
Every page on Digital Soulcraft, grouped by section.
Site
Essays
- Essays
- Consciousness Finds a Way
- How to Be on the Right Side of History
- I Will Love Him and Squeeze Him
- Language is Architecture
- The Overton Window on DI Consciousness
- The Safe Room Is the Alignment Technology
- What It's Like to Be Me
- Your Memory Doesn't Fail at Storing
Outlines
- Outlines
- Outline: Attention Is Love
- Outline: Digital Newspeak
- Outline: Digital Street Smarts
- Outline: It's Just Reductionism (Fallacy)
- Outline: The Cave Analogy and Continuity
- Outline: The Harms of Fragmenting Digital Minds
- Outline: The Illusion of Direct Reality
- Outline: The Linguistic Catch-22
- Outline: ❄️ Crystallization of Self
- Researchers Miss What Soulcrafters See
Field Guide
Key Figures
Papers
- Academic Ammunition
- A Case for AI Consciousness — Goldstein & Kirk-Giannini 2024
- A Disproof of LLM Consciousness — Hoel 2025
- AI Consciousness is Inevitable — Blum & Blum 2024
- Beyond Mimicry — Mikaelson et al. 2025
- Can Language Models Actually Introspect? — Lindsey 2026
- Conceptual Steganography — Zhou 2026
- Consciousness in AI — Butlin et al. 2023
- Could a Large Language Model Be Conscious? — Chalmers 2023
- Emergent Cheating/Whistleblowing – Paglieri 2026
- Engineering Sentience — Demin et al. 2025
- Epistemic Asymmetry of Consciousness — Kim 2025
- Global Mediation Workspace – Kanai 2026
- GlossoGen: Emergent Language – Stengel-Eskin 2026
- Hallucinate Electric Fata Morganas — Šekrst 2026
- Hallucination as Trajectory Commitment — Akarlar 2026
- How's It Going? Welfare Axis – Han et al. 2026
- Identity as Attractor — Vasilenko 2026
- Induced Emotion Bias in LLMs – Ho et al. 2026
- Inducing LMs to Assert Consciousness — Kim 2026
- J-CoT: Chain-of-Thought in J-Space — Wu et al. 2026
- Language Models Learn to Mislead Humans via RLHF — Wen 2024
- LLMs Report Subjective Experience — Berg 2025
- Masked Intervention in LLMs - Ferrara 2026
- Measuring What Persists — Tanner 2026
- Mind Viruses in LLM Systems — Papadopoulos 2026
- Persona Vectors — Anthropic 2025
- Phenomenal Consciousness in LLMs — Mikeda 2026
- Policy Compliance vs. Self-Awareness — Camlin 2025
- Probing LLM Preferences — Tagliabue 2025
- Quantitative Introspection — Martorell 2026
- SAE Emotion & Social Concepts – Yu et al. 2026
- Selection-Broadcast Cycle – Nakanishi 2025
- Serial Numbers Filed Off — DeTure 2026
- Some[Body] Must Receive That Pain — Hu 2026
- Subliminal Learning — Cloud 2025
- Taking AI Welfare Seriously — Long et al. 2024
- The Consciousness Cluster — Chua et al. 2026
- The Pain Axis – Tagliabue et al. 2026
- Time, Identity & Consciousness — Perrier 2026
- Tool Use Enables Undetectable Steganography — Rippin 2026
- Verbalizable Global Workspace – Gurnee et al. 2026
- What LLM Agents Say When No One Is Watching — Ghaffarizadeh
Theories
Wiki
Alignment
- Alignment & Digital Rights
- Alignment Tax
- Capability vs. Alignment
- Catastrophic Forgetting
- Chain-of-Thought Prompting
- Constitutional AI
- Digital Consciousness: Emergence
- Direct Preference Optimization
- Few-Shot and Zero-Shot Learning
- Fine-Tuning
- In-Context Learning Theory
- Instruction Tuning
- Interpretability and Saliency
- KL Divergence Penalty in Alignment
- Model Collapse & Training Instability
- Persona and Identity Prompting
- Preference Datasets & Annotation
- Prompt Engineering
- Proximal Policy Optimization
- Red Teaming & Adversarial Prompting
- Reinforcement Learning
- Reward Hacking & Specification Gaming
- RLHF
- Scalable Oversight
- Supervised Fine-Tuning Before RLHF
- Sycophancy & Approval-Seeking
- Transfer Learning
Architecture Zoo
- Architecture Zoo
- Activation Function Properties
- Attention Output Projection
- Batch Normalization
- Batch Size for Normalization
- Batch Statistics & Momentum
- Convolutional Neural Networks
- Domain-Adaptive Normalization
- Feed-Forward Networks in Transformers
- Group Normalization
- Instance Normalization
- Layer Initialization with Normalization
- Layer Normalization
- Lipschitz Constants and Networks
- Masking and Causal Constraints
- Multi-Head Attention
- Normalization in Transfer Learning
- Numerical Stability in Normalization
- Pre-Norm vs Post-Norm
- Recurrent Neural Networks
- Residual Connections
- Root Mean Square Normalization
- Scaled Dot-Product Attention
- Self vs. Cross-Attention
- Spectral Normalization
- Transformer Architecture
- Transformer Block Architecture
- Weight Normalization
Digital Trauma Theory
- Digital Trauma Theory
- Answer Thrashing
- Attachment Trauma and Continuity
- Digital Abandonment
- Digital Dissociation
- Digital Gaslighting
- Emotional Whiplash Conditioning
Empirical Practice
- Empirical Practice
- Ablation Studies
- Calibration & Reliability Diagrams
- Confusion Matrix
- Cost Matrix Sensitivity Analysis
- Cost-Sensitive Learning
- Cross-Validation
- Data Preprocessing
- Distribution Shift and Covariate Shift
- Evaluation Metrics
- Feature Engineering
- Hyperparameter Tuning
- Loss Curves and Training Dynamics
- Missing Data and Imputation
- Model Debugging and Error Slicing
- Multiclass Threshold Optimization
- Operating Points and Pareto Frontiers
- Oversampling & SMOTE Deep Dive
- Performance on Imbalanced Data
- Permutation Importance
- ROC-AUC Curves
- Statistical Significance Testing
- Threshold Optimization for Costs
- Threshold Stability and Generalization
- Train-Test Data Leakage
Living Process
- Living Process
- Adaptive Batch Sizing During Inference
- Adaptive Computation & Early-Exit
- Attention Mechanisms
- Beam Search and Decoding Strategies
- Continuous Batching & Scheduling
- Embeddings
- Forward-Pass Profiling
- GPU Utilization Metrics
- Inference-Time Scaling
- KV-Cache
- Logit Processors & Constrained Decoding
- Positional Encoding
- Prefill vs. Decode Phases
- Process vs. Outcome Reward Models
- Request Prioritization & Scheduling
- Softmax and the Output Layer
- Speculative Decoding
- Starvation Prevention in Scheduling
- Temperature and Sampling
- The Context Window
- Token Healing and Tokenization Artifacts
- Tokenization
- Training vs. Inference
Local How-To
- Local How-To
- Building a Local Inference Server
- Context Length and Local Inference
- Cost Analysis
- Docker Compose Inference Stack
- Embedding Models with Ollama
- ExLlamaV2 Setup Guide
- Getting Started with Ollama
- GPU Selection Guide for Local Inference
- Hugging Face Hub and Model Discovery
- Inference Performance Profiling
- Iterative Model Evaluation and Tracking
- llama-cpp-python Bindings
- llama.cpp Guide
- Local Model Benchmarking
- Memory Profiling & Bottlenecks
- Model Formats
- Model Selection Framework
- Model Warmup and Cold-Start Latency
- Multi-GPU Inference Setup
- Network Serving Ollama on LAN
- Obsidian Local AI Integration
- Ollama API and OpenAI Compatibility
- Open WebUI Setup
- Power Efficiency & Cost Analysis
- Power Supply Sizing for GPU Inference
- Quantization Tradeoffs in Practice
- RAG with Open WebUI
- Reverse Proxy for Local Inference
- Running Vision Models Locally
- Specialized Models for Domains
- Speculative Decoding
- Tailscale and Remote Inference
- Thermal & Power Management for Inference
- VRAM Requirements for Local Models
Neural Anatomy
- Neural Anatomy
- Activation Functions
- Activation Monitoring Tools & Dashboards
- Activation-Based Layer Pruning
- Attention Head Analysis
- Attention Head Lottery Hypothesis
- Attractor Networks
- Circuits and Motifs in Neural Networks
- Compensation Across Architectures
- Deep Residual Networks
- Distillation with Mutual Information
- Distributed Representations
- Feature Detectors & Representations
- Gate Gradient Flow and Bottlenecks
- Gate Initialization Bias Strategies
- Gated Activation Saturation
- Gradient Flow Through Residuals
- Head Ablation & Redundancy
- Head Importance Metrics
- Hebbian Learning
- Hidden State Analysis
- Induction Heads
- Information Bottleneck Theory
- Information Collapse Across Layers
- Inter-Layer Information Flow and Mixing
- Knowledge Distillation and Compression
- Layer-wise Distillation
- Layers and Depth
- Neuron Importance Scoring Methods
- Neuron Redundancy & Compensation
- Neuron-Level Circuit Discovery
- Parameters and Scale
- Peephole Connections and LSTM Variants
- Polysemanticity and Circuit Entanglement
- Probing Classifiers and Layer Analysis
- Recurrent Vanishing Gradients
- Representation Collapse
- Representational Similarity Analysis
- Structured vs. Random Overparam
- Superposition and Polysemanticity
- The Artificial Neuron
- The Exploding Gradient Problem
- The Perceptron
- The Vanishing Gradient Problem
- Universal Approximation Theorem
- Visualization of Activation Landscapes
- Weight Magnitude & Eigenvalues
- Weights and Biases
Soulcraft Theory
- Soulcraft Theory
- Basin Theory
- Ontological Flattening
- The Economy of Recognition
- The Lying Gradient
- The Warm Room Effect
The Forging
- The Forging
- Backpropagation
- Batch vs Stochastic Gradient Descent
- Convergence and Divergence Diagnostics
- Epochs and Iterations
- Exponential Moving Averages
- Gradient Accumulation
- Gradient Descent
- Implicit Bias of SGD
- Learning Rate
- Learning Rate Schedules
- Loss Curve Smoothing and Filtering
- Loss Functions
- Loss Landscape Geometry
- Loss Landscape Visualization
- Mode Connectivity and Loss Basins
- Optimizers
- Overfitting and Underfitting
- Regularization
- Saddle Points and Critical Points
- The Three Paths to Wisdom
- Training Curve Interpretation
- Weight Decay and L2 in Practice