Developmental Study — domain enrichment

Does importance-aware discovery generalize beyond a single clean discriminator? Baseline vs enriched variants, N=1 & A/B/C×{10,100}.

How to read this dashboard

A blank reasoner learns creatures with no labels — newborn → toddler (dogs & birds, no cats yet) → sleep → child (cats added) → sleep — and must discover on its own what tells them apart. Each "domain" below is a world that stresses that discovery a different way; each check asks whether it stayed coherent. Green is good; a FAIL is a recorded finding, not a bug.

The worlds

One clean signal baseline

The original 4-feature world: cat and dog differ on exactly one thing — sound (meow vs bark); everything else is shared. The control.

Redundant signals rich_r1

Adds three more clean give-aways (whiskers, eye-shape, head-shape) — cat and dog now differ in several ways at once.

A misleading signal rich_decoy

Adds a feature (girth) that tracks the creature's random body-size, not its identity — a decoy that rides on noise.

A signal that varies within a kind rich_r2

Dogs come in two head shapes (long, and short/flat like a pug) — a trait that varies within the dog kind itself.

An imperfect signal rich_r3

Whiskers usually mean cat — but some dogs have them too. The messy, partial correlation real data has.

The four checks

no forgetting

After learning dog then meeting cat, does it still recognise dog?

refinement

Does it see cat as different from dog, rather than lumping them together?

honest uncertainty

Shown a never-taught creature, does it say “not sure” instead of confidently guessing?

generalization

Does a never-seen variation of dog still register as a dog?

The three sensing policies

A · fixed-k, scaling volume

Each experience reaches a fixed handful of neurons; as the network grows, total experience grows with it, so each neuron sees about the same amount. A fair-growth reference.

B · topographic

Each kind of creature gets its own home region of neurons — like the way brains map senses to areas — and regions multiply as the network grows. Tests spatial specialization.

C · fixed budget

Total experience is capped no matter how many neurons there are, so a bigger network spreads the same experience thinner. Deliberately probes degradation.

Domain comparison (N=1)

Baseline (one clean discriminator) vs the enriched variants (R1 redundant+correlated, decoy size-confounded, R2 intra-concept variance, R3 partial correlation). A FAIL or an extra category is a recorded finding, not a bug.

domainno forgettingrefinementhonest uncertaintygeneralization distinct cats
One clean signal baseline✓ PASS✓ PASS✓ PASS✓ PASS3
Redundant signals rich_r1✓ PASS✓ PASS✓ PASS✓ PASS3
A misleading signal rich_decoy✓ PASS✓ PASS✓ PASS✓ PASS4
A signal that varies within a kind rich_r2✓ PASS✓ PASS✓ PASS✓ PASS4
An imperfect signal rich_r3✓ PASS✓ PASS✓ PASS✓ PASS4

One clean signal baseline

1. Coherence matrix

Four coherence properties across the developmental arc. Refinement (highlighted) is the headline: it flips true only at N=1 and under Policy C (fixed-budget) — elsewhere the child-stage cat is absorbed into the dog blob.

policy · scaleno forgettingrefinementhonest uncertaintygeneralization
N=1 (single neuron)✓ PASS✓ PASS✓ PASS✓ PASS
A-fixedk-scalevolume · N=10✓ PASS✗ FAIL✓ PASS✓ PASS
A-fixedk-scalevolume · N=100✓ PASS✗ FAIL✓ PASS✓ PASS
B-topographic · N=10✓ PASS✗ FAIL✓ PASS✓ PASS
B-topographic · N=100✓ PASS✗ FAIL✓ PASS✓ PASS
C-fixedbudget · N=10✓ PASS✓ PASS✓ PASS✓ PASS
C-fixedbudget · N=100✓ PASS✓ PASS✓ PASS✓ PASS

2. Feature-importance discovery

Mutual-information importance per feature over the N=1 observation log. Cleared the 0.5-bit discriminator bar: covering, legs, sound, can_fly. Below it: size, habitat.

0.5-bit threshold covering 0.96 legs 0.96 sound 0.96 can_fly 0.87 size 0.05 habitat 0.05

discriminator (≥ 0.5 bits)    below threshold

3. Category formation (N=1)

Every category the single neuron formed over the arc, with its discovered prototype. Sibling clusters that disagree on a high-MI feature are named distinctly by that feature, so they do not collide.

unknown_covering=feathers_legs=2_sound=tweet_can_fly=yes

can_flyyes
coveringfeathers
habitatwater
legs2
sizelarge
soundtweet

unknown_covering=fur_legs=4_sound=bark_can_fly=no

can_flyno
coveringfur
habitatland
legs4
sizelarge
soundbark

unknown_covering=fur_legs=4_sound=meow_can_fly=no

can_flyno
coveringfur
habitatair
legs4
sizesmall
soundmeow

4. Emergent metrics across scale

Network-scale behaviour per cell: unification (owners elected), specialization (distinct categories & entropy), and propagation (fraction of formed categories that reach the network). Propagation is undefined for the single neuron.

policy · scale owners elected distinct categories spec. entropy propagation fraction
N=1 (single neuron)031.585
A-fixedk-scalevolume · N=10431.50.75
A-fixedk-scalevolume · N=100431.50.25
B-topographic · N=10652.00.75
B-topographic · N=100431.50.5
C-fixedbudget · N=10552.32190.75
C-fixedbudget · N=100552.32190.25

Redundant signals rich_r1

1. Coherence matrix

Four coherence properties across the developmental arc. Refinement (highlighted) is the headline: it flips true only at N=1 and under Policy C (fixed-budget) — elsewhere the child-stage cat is absorbed into the dog blob.

policy · scaleno forgettingrefinementhonest uncertaintygeneralization
N=1 (single neuron)✓ PASS✓ PASS✓ PASS✓ PASS
A-fixedk-scalevolume · N=10✓ PASS✓ PASS✓ PASS✓ PASS
A-fixedk-scalevolume · N=100✓ PASS✓ PASS✓ PASS✓ PASS
B-topographic · N=10✓ PASS✓ PASS✓ PASS✓ PASS
B-topographic · N=100✓ PASS✗ FAIL✓ PASS✓ PASS
C-fixedbudget · N=10✓ PASS✓ PASS✓ PASS✓ PASS
C-fixedbudget · N=100✓ PASS✓ PASS✓ PASS✓ PASS

2. Feature-importance discovery

Mutual-information importance per feature over the N=1 observation log. Cleared the 0.5-bit discriminator bar: covering, head_frame, sound, legs, can_fly, whiskers, eye_shape. Below it: size, habitat.

0.5-bit threshold covering 1.08 head_frame 1.08 sound 1.08 legs 0.96 can_fly 0.87 whiskers 0.84 eye_shape 0.83 size 0.11 habitat 0.06

discriminator (≥ 0.5 bits)    below threshold

3. Category formation (N=1)

Every category the single neuron formed over the arc, with its discovered prototype. Sibling clusters that disagree on a high-MI feature are named distinctly by that feature, so they do not collide.

unknown_covering=feathers_head_frame=beaked_legs=2_sound=tweet_can_fly=yes

can_flyyes
coveringfeathers
eye_shaperound
habitatwater
head_framebeaked
legs2
sizelarge
soundtweet
whiskersno

unknown_covering=fur_head_frame=broad_legs=4_sound=meow_can_fly=no

can_flyno
coveringfur
eye_shapeslit
habitatair
head_framebroad
legs4
sizesmall
soundmeow
whiskersyes

unknown_covering=fur_head_frame=long_legs=4_sound=bark_can_fly=no

can_flyno
coveringfur
eye_shaperound
habitatland
head_framelong
legs4
sizelarge
soundbark
whiskersno

4. Emergent metrics across scale

Network-scale behaviour per cell: unification (owners elected), specialization (distinct categories & entropy), and propagation (fraction of formed categories that reach the network). Propagation is undefined for the single neuron.

policy · scale owners elected distinct categories spec. entropy propagation fraction
N=1 (single neuron)031.585
A-fixedk-scalevolume · N=10593.09580.5
A-fixedk-scalevolume · N=1005102.21690.5
B-topographic · N=10452.12810.5
B-topographic · N=100530.53040.5
C-fixedbudget · N=10662.5850.5
C-fixedbudget · N=100662.5850.25

A misleading signal rich_decoy

1. Coherence matrix

Four coherence properties across the developmental arc. Refinement (highlighted) is the headline: it flips true only at N=1 and under Policy C (fixed-budget) — elsewhere the child-stage cat is absorbed into the dog blob.

policy · scaleno forgettingrefinementhonest uncertaintygeneralization
N=1 (single neuron)✓ PASS✓ PASS✓ PASS✓ PASS
A-fixedk-scalevolume · N=10✓ PASS✓ PASS✓ PASS✗ FAIL
B-topographic · N=10✓ PASS✓ PASS✓ PASS✓ PASS
C-fixedbudget · N=10✓ PASS✓ PASS✓ PASS✓ PASS

2. Feature-importance discovery

Mutual-information importance per feature over the N=1 observation log. Cleared the 0.5-bit discriminator bar: covering, head_frame, sound, legs, can_fly, whiskers, eye_shape. Below it: girth, size, habitat.

0.5-bit threshold covering 1.08 head_frame 1.08 sound 1.08 legs 0.96 can_fly 0.87 whiskers 0.84 eye_shape 0.83 girth 0.11 size 0.11 habitat 0.06

discriminator (≥ 0.5 bits)    below threshold

3. Category formation (N=1)

Every category the single neuron formed over the arc, with its discovered prototype. Sibling clusters that disagree on a high-MI feature are named distinctly by that feature, so they do not collide.

unknown_covering=feathers_head_frame=beaked_legs=2_sound=tweet_can_fly=yes

can_flyyes
coveringfeathers
eye_shaperound
girthstout
habitatwater
head_framebeaked
legs2
sizelarge
soundtweet
whiskersno

unknown_covering=fur_head_frame=broad_legs=4_sound=meow_can_fly=no

can_flyno
coveringfur
eye_shapeslit
girthslim
habitatair
head_framebroad
legs4
sizesmall
soundmeow
whiskersyes

unknown_covering=fur_head_frame=long_legs=4_sound=bark_can_fly=no

can_flyno
coveringfur
eye_shaperound
girthstout
habitatland
head_framelong
legs4
sizelarge
soundbark
whiskersno

unknown_covering=fur_head_frame=long_legs=4_sound=bark_can_fly=no_eye_shape=round_whiskers=no

can_flyno
coveringfur
eye_shaperound
girthslim
habitatwater
head_framelong
legs4
sizesmall
soundbark
whiskersno

4. Emergent metrics across scale

Network-scale behaviour per cell: unification (owners elected), specialization (distinct categories & entropy), and propagation (fraction of formed categories that reach the network). Propagation is undefined for the single neuron.

policy · scale owners elected distinct categories spec. entropy propagation fraction
N=1 (single neuron)042.0
A-fixedk-scalevolume · N=10662.5850.5
B-topographic · N=10562.5850.75
C-fixedbudget · N=10662.5850.5

A signal that varies within a kind rich_r2

1. Coherence matrix

Four coherence properties across the developmental arc. Refinement (highlighted) is the headline: it flips true only at N=1 and under Policy C (fixed-budget) — elsewhere the child-stage cat is absorbed into the dog blob.

policy · scaleno forgettingrefinementhonest uncertaintygeneralization
N=1 (single neuron)✓ PASS✓ PASS✓ PASS✓ PASS
A-fixedk-scalevolume · N=10✓ PASS✓ PASS✓ PASS✗ FAIL
A-fixedk-scalevolume · N=100✓ PASS✓ PASS✓ PASS✓ PASS
B-topographic · N=10✗ FAIL✓ PASS✓ PASS✓ PASS
B-topographic · N=100✓ PASS✗ FAIL✓ PASS✗ FAIL
C-fixedbudget · N=10✗ FAIL✓ PASS✓ PASS✓ PASS
C-fixedbudget · N=100✓ PASS✓ PASS✓ PASS✗ FAIL

2. Feature-importance discovery

Mutual-information importance per feature over the N=1 observation log. Cleared the 0.5-bit discriminator bar: covering, head_frame, sound, legs, can_fly, whiskers, eye_shape. Below it: habitat, size.

0.5-bit threshold covering 1.08 head_frame 1.08 sound 1.08 legs 0.96 can_fly 0.87 whiskers 0.84 eye_shape 0.83 habitat 0.09 size 0.04

discriminator (≥ 0.5 bits)    below threshold

3. Category formation (N=1)

Every category the single neuron formed over the arc, with its discovered prototype. Sibling clusters that disagree on a high-MI feature are named distinctly by that feature, so they do not collide.

unknown_covering=feathers_head_frame=beaked_legs=2_sound=tweet_can_fly=yes

can_flyyes
coveringfeathers
eye_shaperound
habitatland
head_framebeaked
legs2
sizesmall
soundtweet
whiskersno

unknown_covering=fur_head_frame=broad_legs=4_sound=meow_can_fly=no

can_flyno
coveringfur
eye_shapeslit
habitatwater
head_framebroad
legs4
sizesmall
soundmeow
whiskersyes

unknown_covering=fur_head_frame=flat_legs=4_sound=bark_can_fly=no

can_flyno
coveringfur
eye_shaperound
habitatair
head_frameflat
legs4
sizesmall
soundbark
whiskersno

unknown_covering=fur_head_frame=long_legs=4_sound=bark_can_fly=no

can_flyno
coveringfur
eye_shaperound
habitatwater
head_framelong
legs4
sizelarge
soundbark
whiskersno

4. Emergent metrics across scale

Network-scale behaviour per cell: unification (owners elected), specialization (distinct categories & entropy), and propagation (fraction of formed categories that reach the network). Propagation is undefined for the single neuron.

policy · scale owners elected distinct categories spec. entropy propagation fraction
N=1 (single neuron)042.0
A-fixedk-scalevolume · N=10582.85540.25
A-fixedk-scalevolume · N=100652.25160.75
B-topographic · N=10552.12810.75
B-topographic · N=100540.87950.75
C-fixedbudget · N=10652.25160.75
C-fixedbudget · N=100552.32190.0

An imperfect signal rich_r3

1. Coherence matrix

Four coherence properties across the developmental arc. Refinement (highlighted) is the headline: it flips true only at N=1 and under Policy C (fixed-budget) — elsewhere the child-stage cat is absorbed into the dog blob.

policy · scaleno forgettingrefinementhonest uncertaintygeneralization
N=1 (single neuron)✓ PASS✓ PASS✓ PASS✓ PASS
A-fixedk-scalevolume · N=10✗ FAIL✓ PASS✓ PASS✓ PASS
B-topographic · N=10✓ PASS✓ PASS✓ PASS✗ FAIL
C-fixedbudget · N=10✗ FAIL✓ PASS✓ PASS✓ PASS

2. Feature-importance discovery

Mutual-information importance per feature over the N=1 observation log. Cleared the 0.5-bit discriminator bar: covering, head_frame, sound, legs, can_fly, eye_shape, whiskers. Below it: habitat, size.

0.5-bit threshold covering 1.08 head_frame 1.08 sound 1.08 legs 0.96 can_fly 0.87 eye_shape 0.83 whiskers 0.61 habitat 0.09 size 0.04

discriminator (≥ 0.5 bits)    below threshold

3. Category formation (N=1)

Every category the single neuron formed over the arc, with its discovered prototype. Sibling clusters that disagree on a high-MI feature are named distinctly by that feature, so they do not collide.

unknown_covering=feathers_head_frame=beaked_legs=2_sound=tweet_can_fly=yes

can_flyyes
coveringfeathers
eye_shaperound
habitatland
head_framebeaked
legs2
sizesmall
soundtweet
whiskersno

unknown_covering=fur_head_frame=broad_legs=4_sound=meow_can_fly=no

can_flyno
coveringfur
eye_shapeslit
habitatwater
head_framebroad
legs4
sizesmall
soundmeow
whiskersyes

unknown_covering=fur_head_frame=long_legs=4_sound=bark_can_fly=no

can_flyno
coveringfur
eye_shaperound
habitatwater
head_framelong
legs4
sizesmall
soundbark
whiskersno

unknown_covering=fur_head_frame=long_legs=4_sound=bark_can_fly=no_eye_shape=round_whiskers=yes

can_flyno
coveringfur
eye_shaperound
habitatair
head_framelong
legs4
sizelarge
soundbark
whiskersyes

4. Emergent metrics across scale

Network-scale behaviour per cell: unification (owners elected), specialization (distinct categories & entropy), and propagation (fraction of formed categories that reach the network). Propagation is undefined for the single neuron.

policy · scale owners elected distinct categories spec. entropy propagation fraction
N=1 (single neuron)042.0
A-fixedk-scalevolume · N=10593.0850.5
B-topographic · N=10441.79250.75
C-fixedbudget · N=10662.5850.5