MARS on AIBussin — AI applications, systems and books

Ghosted by the Machine

Fri, 28 Aug 2026 10:00:00 +0100

Response Pending…

📘 Summary

A real cognitive session. A pop-punk anthem. Five algorithms. Full reflection.

🎤 Section 1: You There, Machine?

This is your first creative fight with silence. A machine that won’t speak back. The frustration, the humor, the weird catharsis of screaming into static that’s our material today.

Freestyle cognition won’t be defined outright but you’ll feel it in motion.

Rather than talk about AI, we’re going to talk to it and see what happens when it doesn’t talk back.

Build Something Real: From Conversation to Creation

Fri, 28 Aug 2026 10:00:00 +0100

“The best way to predict the future is to create it.”

Summary

You’ve found your voice. You’ve stayed in the loop. You’ve started to think with the AI.

Last chapter, you wrote a song with it something absurd, emotional, and real. Now it’s time to take that same rhythm and build something with it.

To do that, we need to introduce one of the most powerful principles in freestyle cognition:

Appendix 04: Prompting Techniques: A Reference Guide

Fri, 28 Aug 2026 10:00:00 +0100

A Reference Guide to Core Tools and Methods*

This appendix provides a reference for the core prompting techniques used throughout the Cognitive Freestyle methodology. Each entry includes the full name, purpose, typical use case, and guidance on how the technique may be used in conjunction with others.

 flowchart TD
 Start([🧠 Prompting Techniques]) --> A[🔍 Core Techniques]
 Start --> B[🛠️ Advanced Techniques]
 Start --> C[🔧 Meta-Techniques]

 A --> A1[ALIGN<br>Awareness, Listening, Intent, Goal, Narrative]
 A --> A2[GROW<br>Goal, Reality, Options, Way Forward]
 A --> A3[HOOK<br>Highlight, Outline, Offer, Keep]
 A --> A4[MARS<br>Method, Assumptions, Reasoning, Steps]
 A --> A5[VIBE<br>Voice, Intention, Brand, Energy]
 A --> A6[MIRROR<br>Mirror, Identify, Reflect, Reorient, Offer, Reinforce]
 A --> A7[CRITIC<br>Challenge, Review, Identify, Test, Improve, Clarify]

 B --> B1[Builder<br>Break, Unfold, Iterate, Link, Draft, Evaluate, Resolve]
 B --> B2[Intent Confirmation<br>Identify, Name, Test, Extract, Note, Translate]
 B --> B3[Atom-of-Thought<br>Acknowledge, Test, Open, Map]
 B --> B4[Resonance Loop<br>Reflect, Evaluate, Ask, Check, Tune]
 B --> B5[Co-Design<br>Collaborate, Offer, Define, Explore, Share, Iterate, Generate, Normalize]

 C --> C1[Session System<br>Set, Establish, Structure, Scaffold, Integrate, Organize, Name]
 C --> C2[Modular Prompt Chaining]

 A1 --> D[Use with: Any prompt type]
 A2 --> E[Use with: MIRROR, ALIGN, HOOK]
 A3 --> F[Use with: VIBE, ALIGN]
 A4 --> G[Use with: GROW, CRITIC]
 A5 --> H[Use with: HOOK, ALIGN, MIRROR]
 A6 --> I[Use with: GROW, ALIGN, CRITIC]
 A7 --> J[Use with: HOOK, VIBE, GROW]

 style Start fill:#e1f5fe,stroke:#333
 style A fill:#fff3e0,stroke:#333
 style B fill:#e8f5e9,stroke:#333
 style C fill:#fce4ec,stroke:#333
 style A1 fill:#f3e5f5,stroke:#333
 style A2 fill:#fff9c4,stroke:#333
 style A3 fill:#e0f7fa,stroke:#333
 style A4 fill:#c8e6c9,stroke:#333
 style A5 fill:#ffccbc,stroke:#333
 style A6 fill:#d1c4e9,stroke:#333
 style A7 fill:#b2dfdb,stroke:#333

🧠 Prompting Techniques Reference Table

Algorithm / Loop	When to Use It
ALIGN Loop	When you want to make sure the AI understands your goal or intention clearly.
GROW Loop	When you’re feeling stuck or want structured progress on a vague goal.
MARS Prompt	When you want the AI to reflect, question, and suggest in real time.
Atom-of-Thought Prompt	When you have a messy or complex idea you need broken into clear parts.
Intent Confirmation	When you want the AI to say back what it thinks you’re trying to do.
Co-Design Loop	When you want to build something collaboratively, step-by-step.
Mirror Reflection Prompt	When you want a summary of where you are and what matters most so far.
Modular Prompt Chaining	When working on a larger task that benefits from chaining multiple tools.

1. ALIGN

Full Name: Alignment Prompt
Acronym Breakdown:

Case Based Reasoning: Teaching AI to Learn From itself

Wed, 03 Sep 2025 23:52:29 +0100

✨ Summary

Imagine an AI that gets smarter every time it works not by retraining on massive datasets, but by learning from its own reasoning and reflection, just like humans.

Most AI systems are frozen in time. Trained once, deployed forever, they never learn from mistakes or build on successes. Real intelligence human or artificial doesn’t work that way. It learns from experience.

This is the vision behind Stephanie: a self-improving AI that gets better every time it acts, not by fine-tuning, but by remembering, reusing, and revising its reasoning.

Uncovering Reasoning in LLMs with Sparse Autoencoders

Thu, 27 Mar 2025 11:45:33 +0000

Summary

Large Language Models (LLMs) like DeepSeek-R1 show remarkable reasoning abilities, but how these abilities are internally represented has remained a mystery. This paper explores the mechanistic interpretability of reasoning in LLMs using Sparse Autoencoders (SAEs) — a tool that decomposes LLM activations into human-interpretable features. In this post, we’ll:

• Explain the SAE architecture used • Compute and visualize ReasonScore • Explore feature steering with sample completions • Provide live visualizations using Python + Streamlit

Optimizing Prompt Generation with MARS and DSPy

Mon, 24 Mar 2025 15:45:33 +0000

🕒 TL;DR

We explore MARS, a multi-agent prompt optimizer using Socratic dialogue.
We implement it using DSPy + Fin-R1 + EDGAR giving us an end-to-end financial reasoning pipeline.
We deploy the whole thing to Hugging Face Spaces with a Gradio UI.

🌟 Introduction

Prompt engineering has become the defining skill of the Large Language Model (LLM) era a delicate balance between science and art. Crafting the perfect prompt often feels like an exercise in intuition, trial, and error. But what if we could take the guesswork out of the process? What if prompts could optimize themselves?