What changed
The raw comparison names 80,000 possible family/substrate sites, binary persistence, continuous Ψ–κ activation, verified-confidence gating and deficit-based service fairness. This is a material architectural change from storing one 16-value vector per agent.
Learning path
The master increment multiplies remaining capacity (1−φ), learning drive Γ and effective allocation π_eff. The drive multiplies activation, bounded plasticity and confidence. The allocator combines match, grounded confidence, novelty, existing coherence and a power-sharpened distribution. A high-level description of these factors is not enough to establish the source’s broad no-hallucination guarantee.
Run and persistence evidence
The archived session reports a 5,000-tick run with 13 learn events and a binary header active count of 26. The log and snapshot paths are historical locators. The newly opened archived SQLite database has 35,338 event rows but only 65 distinct item texts; repeated event counts therefore cannot be equated with distinct knowledge acquisitions.
Equations in this build
41 linked mathematical records. Each opens its notation, interpretation and source.
Golden-frequency sine–cosine feature
The diagram defines a feature as the sum of sine and cosine at the same golden-scaled phase. The cosine continuation was on the next diagram line and is included here. This gives…
Learning · SEFERIM FEDERATION - ARCHITECTURAL DIAGRAM · θ_k +
The expression assigns a hash-derived representation to the listed state components. A hash is a many-to-one identifier unless a reversible representation is explicitly specified;…
Query · SEFERIM FEDERATION - ARCHITECTURAL DIAGRAM · K
The expression assigns a hash-derived representation to the listed state components. A hash is a many-to-one identifier unless a reversible representation is explicitly specified;…
Cosine coherence between knowledge vectors
Computes the normalized dot product of two knowledge vectors. It compares vector direction and is undefined if either vector has zero norm; a practical implementation needs a…
Decay-and-coupling federation dynamics
Combines a decay term −λS_i with a weighted sum of other agents’ tanh-transformed signals. Coherence weighting changes the effective coupling. Stability depends on the decay rate,…
Hop-by-hop federation activation propagation
Adds neighboring activation contributions to an agent’s previous-hop activation, weighted by graph edges and coherence. Repeated hops can amplify signals unless the network…
Multiplicative learning drive
Multiplies activation χ, plasticity η and confidence κ. Any zero factor disables the drive, while each positive factor scales the contribution of the others.
Activation · Mathematical source · χ
The expression uses a saturating nonlinear transform to turn its input combination into an activation or bounded score. The coefficients control sensitivity and the offset…
Confidence-controlled plasticity
Adds a confidence-dependent gain to the base learning rate, then clips it between η_min and η_max. Ordered finite bounds and a well-defined confidence value are required.
Pressure normalization in code
Computes pressure divided by pressure plus a reference scale. For nonnegative pressure and a positive scale, the result is bounded below one.
EXTERNAL WORLD SEFERIM CORE PERSISTENCE · float chi
The expression uses a saturating nonlinear transform to turn its input combination into an activation or bounded score. The coefficients control sensitivity and the offset…
Clipped plasticity parameter
Forms a confidence-dependent learning rate and clips it between the configured minimum and maximum. The bounds control allowed update intensity.
Learning-drive product in code
Computes the implementation’s product of activation, learning rate and item confidence, matching the proposed Γ factor.
CONTROLLER │◀────────▶│ FEDERATION │◀────────▶│ ALLOCATOR │ │ BINARY I/O
This relation declares the set in which the stated variable is allowed to take values. It is a domain constraint or type declaration, not a state-update algorithm.
Implementation of saturating substrate learning
Computes an increment proportional to remaining coherence capacity and learning drive, then stores the incremented value with an upper clip at one. Nonnegative increments require…
LEARNING FUNCTION · π²_s
The expression uses a saturating nonlinear transform to turn its input combination into an activation or bounded score. The coefficients control sensitivity and the offset…
Grounded verification confidence κ real
κ real is the source’s verification-confidence factor in the effective substrate-allocation rule. Its intended purpose is to suppress updates from unverified content. The claimed…
Power-sharpened allocation distribution
Raises each nonnegative allocation to power γ and renormalizes. For γ>1 this concentrates relative mass on larger entries. The denominator must remain positive; an all-zero vector…
STAGE 4: COMPETITION SHARPENING + NORMALIZE · learn
The expression uses a saturating nonlinear transform to turn its input combination into an activation or bounded score. The coefficients control sensitivity and the offset…
STAGE 4: COMPETITION SHARPENING + NORMALIZE · 2. Compute activation: χ
The expression uses a saturating nonlinear transform to turn its input combination into an activation or bounded score. The coefficients control sensitivity and the offset…
Clipped plasticity parameter
Forms a confidence-dependent learning rate and clips it between the configured minimum and maximum. The bounds control allowed update intensity.
Learning-drive product
Scales learning by the product of activation, plasticity and confidence. The product couples readiness to learn with rate and evidence strength.
STAGE 4: COMPETITION SHARPENING + NORMALIZE · SEFERIM_ALPHA_KAPPA
This statement fixes a parameter value for the documented model version. Its effect depends on the equation or update rule that consumes that parameter; it is retained separately…
STAGE 4: COMPETITION SHARPENING + NORMALIZE · SEFERIM_ALPHA_PSI
This statement fixes a parameter value for the documented model version. Its effect depends on the equation or update rule that consumes that parameter; it is retained separately…
STAGE 4: COMPETITION SHARPENING + NORMALIZE · SEFERIM_THETA
This statement fixes a parameter value for the documented model version. Its effect depends on the equation or update rule that consumes that parameter; it is retained separately…
STAGE 4: COMPETITION SHARPENING + NORMALIZE · SEFERIM_PSI_0
This statement fixes a parameter value for the documented model version. Its effect depends on the equation or update rule that consumes that parameter; it is retained separately…
STAGE 4: COMPETITION SHARPENING + NORMALIZE · SEFERIM_ETA_MIN
This statement fixes a parameter value for the documented model version. Its effect depends on the equation or update rule that consumes that parameter; it is retained separately…
STAGE 4: COMPETITION SHARPENING + NORMALIZE · SEFERIM_ETA_MAX
This statement fixes a parameter value for the documented model version. Its effect depends on the equation or update rule that consumes that parameter; it is retained separately…
STAGE 4: COMPETITION SHARPENING + NORMALIZE · SEFERIM_ETA_BASE
This statement fixes a parameter value for the documented model version. Its effect depends on the equation or update rule that consumes that parameter; it is retained separately…
STAGE 4: COMPETITION SHARPENING + NORMALIZE · SEFERIM_ETA_GAIN
This statement fixes a parameter value for the documented model version. Its effect depends on the equation or update rule that consumes that parameter; it is retained separately…
Illustrative activation regime
The notes describe an expected activation level for this qualitative pressure/confidence regime. The displayed approximation is illustrative because numeric inputs for “high” and…
Illustrative activation regime
The notes describe an expected activation level for this qualitative pressure/confidence regime. The displayed approximation is illustrative because numeric inputs for “high” and…
Illustrative activation regime
The notes describe an expected activation level for this qualitative pressure/confidence regime. The displayed approximation is illustrative because numeric inputs for “high” and…
Illustrative activation regime
The notes describe an expected activation level for this qualitative pressure/confidence regime. The displayed approximation is illustrative because numeric inputs for “high” and…
Fairness service deficit
Computes target share times total service events minus actual services for family f. A positive deficit means the family has received less service than its target share.
Largest-deficit family selector
Chooses a family with maximal service deficit. This implements a fairness priority when the deficit definition and service accounting are consistent; ties require a selection…
Strongest shared-pool confidence
Takes the maximum reliability-weighted match score across pool candidates. The aggregation chooses the best available candidate according to R_k rather than combining independent…
Bounded shared-pool contribution
Scales the difference between pool and local confidence by β_max and clamps the contribution. The source’s clamp argument order is inconsistent with common signatures, so the…
Ignore weaker shared evidence
States the intended policy that a pool whose confidence does not exceed the local confidence contributes zero. It is a boundary case of the proposed bounded pool-sharing rule.
Source symbol definition
This line gives the source-local meaning of the displayed symbol. It is a definition used by surrounding equations; the same Greek letter can have a different meaning in another…
Saturating substrate learning update
Increments substrate coherence by remaining capacity (1−φ) times learning drive and effective allocation. If φ, Γ and allocation are all in [0,1], the update stays in [0,1] and…