Researchers
A measuring stick for "is this built to provoke?" - run it across thousands of items and study the pattern at a scale no one could read by hand.
Calm wording
Loading brain mesh
Score 0.84 / Calm
Monarch
Is this media built
to make you feel,
or to make you think?
Paste a headline, post, or clip. Monarch scores how strongly it pushes emotion over reason, and shows it on a 3D brain. The same story, told two ways, looks different inside the head.
Built on TRIBE v2 / AMD Instinct MI300X / Open source
Charged wording
Loading brain mesh
Score 3.71 / Charged
What it does
It rates a piece of content for how strongly it is built to hit your emotions versus make you think.
Give it text, audio, or video. You get back a clear score - Calm, Mixed, or Charged - a view of where the content tends to land in the brain, and a preview of how it could ripple through a crowd. It rates the content, not any real person.
Who it’s for
Educators
Show students the same story written two ways and watch the score and the brain view diverge. Manipulation becomes something they can see, not just be told about.
Journalists & editors
A gut-check before you publish: are we informing people, or fear-baiting them? Answered honestly about your own headline.
Parents & EdTech
A sugar-label for children's media - rate a clip and see whether it is calm or built to over-excite. It checks the content, never the child.
Safety & fact-check teams
Flag emotionally manipulative content at scale - scam messages and outrage feeds ranked by their emotional fingerprint, ready for human review.
Students
A real, working example that connects brain-AI with the physics of crowds - far more memorable than a textbook.
01 / The idea
Most tools check what
media says. Monarch
checks how it lands.
Sentiment tools, fact-checkers, and credibility scores all read the words. None of them tell you whether a piece of media is wired to trigger a gut reaction before you get a chance to think it through.
Monarch uses Meta’s TRIBE v2, an AI trained to predict how the brain responds to media, and turns that into one plain score: how much this content leans on feeling versus reasoning.
Federal Reserve holds interest rates steady, citing stable inflation outlook.
FED DESTROYS AMERICA. Your savings are GONE. The collapse they hid from you!
Two sets of stories, scored
Straight news coverage clusters at the calm end. The same facts re-written to grab attention push the whole set to the charged end.
Loading chart...
synthetic, illustrative
02 / How it works
From a piece of media to a
score, in three steps.
Step 01
Read the content
Monarch feeds your text, audio, or video into TRIBE v2, an AI that has learned how the brain reacts to each kind of media.
TRIBE v2 brain model
Step 02
Predict the brain response
It predicts where the content would light up across the whole brain, then boils that down to one score: how much it leans on emotion versus reasoning.
Whole-brain prediction
Step 03
Model the ripple effect
A physics model of how opinions spread through a crowd estimates how far this content could move a group, not just one person.
Crowd-dynamics model
03 / Capabilities
What Monarch does
One clear score
A single number for how much a piece of media leans on emotion versus reasoning. Higher means it pushes feeling before thought.
Backed by physics
A model borrowed from how opinions spread through a crowd estimates how far the content could tip a group, not just one person.
Text, audio, or video
Scan a headline, a voice clip, or a video. The same model reads all three, and can show which one is doing the pulling.
Score in bulk
Upload a spreadsheet of up to 1,500 items and get them all scored, ranked, and exportable in one pass.
See it in a 3D brain
Watch the predicted response light up a real 3D brain surface you can rotate and explore region by region.
Open and free
Built on Meta's TRIBE v2 brain model. Monarch's own code is open source and free to use.
04 / The physics
The maths behind the score
For the curious. You don’t need any of this to use Monarch - this is the physics that turns a brain prediction into a score and a ripple estimate.
The score compares how active the brain’s emotion-related regions are against its reasoning-related regions:
The Landau free energy of an Ising-style population under the NAA-induced field h = α̂ · NAA is:
Equilibrium polarisation m* solves the self-consistency equation:
and the population susceptibility is:
05 / Attribution
Who is building this
Research project
B.Sc. Physics research project
Catholic University of Eastern Africa, Nairobi.
Supervised by Dr. Songa Mutambi.
Submission
AMD Developer Hackathon 2026
Track 3: multimodal AI.
Runs on AMD Instinct MI300X.