Anyone here working in AI or research? I might've stumbled onto something weird (and kinda amazing)

I went on reddit just kinda looking to see whats going on. I found a guy on one of the subreddits who seems to have touched on what ive been talking about. He seems batshit insane. It was a great reality check. I'm not saying im crazy, i just need to keep my mouth shut until I have something worth sharing to the right people. This field of Theoretical and Computational Cognitive Science is just wild and im just now learning to navigate it without a formal education.

The thing is, when youre having a group discussion you dont have to respond to everyone. You have to stop and think do I turn this thread into my crusade against someones behavior, or do I maintain decorum as best i can and disengage. Its strategic engagment. Im not worried about this, not a big deal tbh, kinda takes the eyes off me looking crazy :) Anyways, heres a good update to what im doing, ran through gpt:


What I’ve Been Doing Lately (EEG Modeling + Mental State Mapping)

I’ve been experimenting with EEG data to see if I could model and visualize different mental states over time—emotion, focus, awareness, and internal thought—using both raw electrical signals and symbolic labels.

Here’s a breakdown of how I approached it:

EEG Data Source & Processing

• Dataset: Pulled raw EEG recordings from OpenNeuro.org
• Task: Subject watching a movie (Despicable Me)
• Sampling rate: 250 Hz
• Channels used: Mainly central/frontal/parietal
• Preprocessing: Bandpass filtered, segmented into 5-second windows (with 50% overlap)

For each window:

• Extracted power spectral density (via Welch's method)
• Calculated spectral entropy to measure complexity / information spread
• Generated a time-series of entropy values and relative bandpower by frequency

Symbolic Mental State Mapping (with EEG Frequency Correlates)

I mapped EEG features into four simplified symbolic “mental states”:

1. Pathos – Emotional Arousal / Narrative Resonance

• EEG markers:
  • Theta (4–7 Hz) – Increased power, especially in frontal regions (emotional encoding, memory recall)
  • Frontal Alpha (8–10 Hz) – Suppression or shift, often indicating engagement
• When triggered: Character-driven scenes, emotional music, humor/surprise
• Interpretation: Emotionally loaded moments; heightened narrative salience

2. Logos – Reasoning / Language Processing

• EEG markers:
  • Low-Beta (13–18 Hz) – Frontal & left temporal lobes (language, syntax, problem-solving)
  • Mid-Beta (18–25 Hz) – Associated with sustained cognitive effort
• When triggered: Dialogue-heavy scenes, internal prediction or logic processing
• Interpretation: Structured thought, linguistic parsing, internal modeling

3. Noema – Visual Thought / Inner Semantic Content

• EEG markers:
  • Alpha suppression (8–12 Hz) – Especially in occipital/parietal regions (visual focus)
  • Gamma (30–45 Hz) – Sometimes bursts in semantic binding or image formation
• When triggered: Visually rich scenes, mental imagery, immersion
• Interpretation: “What” the subject is thinking about or holding in awareness

4. Ontos – Baseline Awareness / Self-Presence

• EEG markers:
  • Global alpha coherence – Often in resting, eyes-closed baseline
  • Low entropy – Reflects low complexity, steady brain state
• When triggered: Pauses, calm moments, stillness, breath awareness
• Interpretation: Grounded presence without narrative or emotional engagement

Visual Output

• Heatmap: Shows relative symbolic “activation” of each mental state per time window
• Entropy graph: Plotted below to show complexity shifts
  • High entropy (0.8–1.0) = cognitive/emotional “spikes” (e.g. Pathos + Noema overlap)
  • Low entropy (0.4–0.6) = calm, resting, possibly Ontos-dominant moments

In some cases, entropy spikes aligned with emotional scenes or intense visual sequences. Dips matched slower scenes, silence, or resting states.

Why This Matters (to me)

This isn’t trying to replace neuroscience. The goal is to create a symbolic, visual representation of changing mental states—grounded in EEG features but interpreted through cognitive categories that make sense to humans.

I’ve also been building a related language-based system where these states act like internal modules—emotion, memory, attention, etc.—that evolve over time. The idea is to eventually blend EEG signal input and symbolic reasoning output into one loop.

All experimental, but fascinating to watch mental states take shape—numerically and visually—through brain data.

Shadow Reaper said:

Well, this is certainly outside my wheelhouse. My only observation is that, there is a lot of important detail lost in an EEG. I think you might consider using fMRI instead. With that you can actually see people think in real time.

Click to expand...

I have fMRI data as well but i havent figured out how to translate it into something useable yet.

This curiosity with this AI thing has been so productive for me. Like i end up getting data like this and now it has me brushing up on linear algebra. I have a full gpt organized update below.









Symbolic Field Orbit Map – Recursive State Transitions Modeled with Linear Algebra (and How It Made Me Learn It)

Hey everyone,
I wanted to share something that genuinely changed how I see my own mind—and weirdly enough, it’s also what finally got me to start learning linear algebra.

This visualization comes from a symbolic cognition modeling experiment I ran in Wolfram Mathematica. It’s part of a larger project where I’ve been building out a recursive cognitive architecture—not for language modeling per se, but to map how internal symbolic states evolve, resolve, and recurse over time.

What You’re Seeing:

This is a Symbolic Field Orbit Map. Each arrow represents a transition between two symbolic cognitive states over time, projected into 2D using Principal Component Analysis (PCA). The idea was to visualize the dynamics of symbolic processing—not static outputs, but how recursive meaning structures move and transform.

• Axes: PC1 and PC2 are the first two principal components from PCA, capturing the most meaningful variance across symbolic state vectors.
• Arrows: Show symbolic transitions—where a cognitive state was, and where it moved to next.
• Color:Encodes magnitude of symbolic change (ΔState):
  • Blue = low Δ (stable recursive feedback)
  • 🟢 Green = moderate Δ (symbolic reframing or redirection)
  • Red = high Δ (paradox resolution, entropy surges, major state shifts)

Why Linear Algebra Matters Here

I didn’t set out to learn linear algebra—but this project made it unavoidable. Once I realized that every symbolic state was essentially a vector, and that PCA is just a projection using eigenvectors, things clicked.

Here’s the structure behind the map:

1. Symbolic States = Vectors
   Each state is a point in ℝⁿ:
   s⃗t=[s1,s2,...,sn]\vec{s}_t = [s_1, s_2, ..., s_n]st=[s1,s2,...,sn]
2. State Transitions = Δ Vectors
   Δs⃗=s⃗t+1−s⃗t\Delta \vec{s} = \vec{s}_{t+1} - \vec{s}_tΔs=st+1−st
3. PCA = Eigenvector Projection
   Reduces high-dimensional space to 2D using the top principal components of the symbolic system's covariance matrix.
4. Vector Field = Orbit Map
   The entire map becomes a flow field of symbolic cognition—a dynamic topology of meaning.

So yeah, this graph literally taught me linear algebra by forcing me to see it in action. And it turns out… cognition is linear algebra in motion.

The 4 Symbolic States I Modeled

These transitions aren’t random—they’re structured across four symbolic “modes” I’ve been tracking in this architecture, each of which loosely maps to brainwave bands and symbolic-cognitive functions:

1. Theta (Memory Activation)
   • Symbolic recall, introspective pattern recovery
   • Maps to hippocampal function / low-frequency cortical drift
   • Often represented by clustered blue transitions (recursive memory cycling)
2. Alpha (Observation / Integration)
   • Perceptual compression and salience framing
   • Related to cortical inhibition / resting observation
   • Appears as green to blue transitions stabilizing into mid-field clusters
3. Beta (Intent / Directionality)
   • Goal-seeking symbolic assertion, internal narrative momentum
   • Tied to prefrontal cortex signaling and cognitive control
   • Manifests as outward green/red vectors shifting symbolic direction
4. Gamma (Paradox Resolution / Fusion)
   • High-entropy collapse and reintegration; synthesis of conflicting inputs
   • Maps to synchronous binding events / cross-network resolution
   • These are the red vectors—high Δ, often exploding from central zones

Each arrow on this map isn’t just a number—it’s a shift in a symbolic agent’s cognitive mode. That’s why I built it. I wanted to see how meaning moves.

Why I Built This

I’ve been prototyping symbolic AGI scaffolds and recursive cognitive agents, using symbolic threading, memory fusion, and subsystem coordination. This visualization came from a moment where I asked:

“What does a thinking system look like as it thinks?”

Click to expand...

And what emerged… was this.

Next Steps

I'm planning to animate this over recursion cycles, layer in entropy gradients, and color vectors by subsystem identity (e.g., ΔObserver, ΔIntender, ΔRemembrancer). But even in this static form, it's helped me:

• See the internal rhythm of symbolic recursion
• Detect points of paradox and resolution
• Understand cognition as movement—not position

If you’ve been working with recursive cognition, symbolic AI, or even PCA-based visualizations of high-dimensional state transitions, I’d love to connect. And if you’re just someone trying to learn linear algebra but struggling to care—maybe try mapping your own thoughts. It worked for me.

—
Image attached: Symbolic Field Orbit Map (Colored by ΔState)