Read every agent state through a real fly brain
Any state in, 16 signed compartment numbers and one scalar out.
No dependencies. No network at runtime. About 0.65 ms per read.
npx skills add https://github.com/valencefly/valence
How it reads
Four stages, three real synapse tables.
- 01. Project
- 02. Expand
- 03. Sparsify
- 04. Read out
Signed by the fly
The sign of each compartment is the sign of its own PAM to PPL1 presynapse ratio.
Derived signs
The ratio decides. Nobody chose them.
No update rule
Same input, same output, forever.
Zero dependencies
One Node process. No network.
Proven red
18 gate arms, each broken once on purpose.
The compartments
Sixteen chips: the name, and the sign its dopamine input gives it.
See every compartment- CA0
- a'1-
- a'2-
- a'3-
- a1+
- a2-
- a3-
- b'1+
- b'2+
- b1+
- b2+
- g1-
- g2-
- g3+
- g4+
- g5+
+ appetitive, - aversive, 0 no teaching signal.
Give your agent a gut
Install it, read a state, compare the scalar across states.
- Install it
- Read a state
- Compare the scalar
Install valence
MIT licensed. Nothing to train.