Papers · Atlas · Data

The history space

Research note

Richard Hoekstra · April 2026

The one-line claim

Thesis The Platonic object is not the compressor, not the swarm, not the Hamiltonian, and not the state space. It is the weighted space of admissible histories with boundary conditions. Everything else is a chart on it.

Formal definition

The history space is the tuple:

h = (Γ_adm, ∂_in, ∂_out, A, C)

Component	Definition
Γ_adm	The set (or groupoid) of admissible histories: sequences of steps through a guarded bulk, consistent with dynamics U, R and every applicable Hoare-style corridor
∂_in : Γ_adm → Σ*	Incoming boundary projection
∂_out : Γ_adm → Σ*	Outgoing boundary projection
A : Γ_adm → Q_≥0	Action functional (local costs summed along the history)
C	Cochain / topological data: exact / harmonic / residual decomposition, charge, current, potential

The Lean symbolic side is already in Scattering.lean: History X Sym U E, WellFormed D γ, Admissible D g γ, inboundary γ, outboundary γ, action L γ with action_nil, action_cons, action_append.

The generating function

The partition function over admissible histories with fixed boundary conditions:

Z(λ; bⁱⁿ, b^out) = Σ_{γ ∈ Γ_adm}
    exp(−λ · A(γ))
    1[∂_in γ = bⁱⁿ]
    1[∂_out γ = b^out]

At λ = 1: the unnormalised scattering weight (already proven correct for the deterministic codec case). As a function of λ: a spectral object whose asymptotics encode entropy rate, whose poles encode phase transitions, and whose derivatives encode expected action.

When Γ_adm collapses to a single admissible path (a deterministic codec with fixed boundaries), Z is a single exp(−A). When guards allow branching — nondeterministic routing, stochastic corrections, MCTS lookahead — the sum has real content.

Every artefact is a chart

Artefact	Chart on h	What it exposes
tlc_compressor.py	Single deterministic γ	Boundary coding of one history
dual_compressor.py	Same γ in Q[ε]/(ε²)	Multiplicative action form
casimir_compressor.py	Families of γ with different guard structure	ΔL as comparison of two charts
atlas_spectrum.py	Ladder of FiniteFamily + advantage curves	Measured projection of h onto rungs
hyperbolic_compressor.py	Local chart where branching is geometrically cheap	Hyperbolic coordinates
open_kernel.py	EncoderKernel / DecoderKernel	Same γ with swapped ports
scattering.py	Enumerator + Σ exp(−A) over finite slice of Γ_adm	Sum-over-histories directly
swarm_model.py	Measure on a local chart of Γ_adm	Bayesian filter over charts
guarded_mcts.py	PUCT exploration of a subtree of Γ_adm	Search within the history space
thermal_swarm.py	Z(λ) at various λ	Finite-temperature slices
quantum_swarm.py	Replace Σ exp with \|Σ e^iΦ\|²	Born rule version of Z
solomonoff_swarm.py	2^−L prior over generators of histories	Universal-prior sum

The Ramanujan program

Stop asking "what is the next state?" and ask "how does the space of admissible histories count?" Four questions become well-posed:

Q1: Dirichlet series D(s; bⁱⁿ, b^out) = Σ_γ A(γ)^−s. Well-defined when A(γ) has discrete values. Its abscissa of convergence encodes the growth rate of admissible histories by action. For a codec on a bounded alphabet this is the byte-conditional entropy rate.

Q2: Euler product If guards factor through local-at-each-step constraints, Z factorises as Π_step Σ_{u, ε valid} exp(−L(·)). Each factor is a finite sum. This is the Euler side of the history-zeta.

Q3: Encode-decode duality The port-swap (relabel boundary in ↔ out) is a Z₂ symmetry: Z(λ; bⁱⁿ, b^out) → Z(λ; b^out, bⁱⁿ). Encoder and decoder are charts of each other.

Q4: Poles of Z(λ) A pole in Z(λ) at some λ_c means a phase transition of the computation — a regime switch. On a regime-mixed corpus, Z(λ) should show real critical structure.

The formal summary

Concept	In the history space
Computation	Choice or sum over γ ∈ Γ_adm
Execution	Min-action γ ∈ Γ_adm
Generation	Sample from e^−A / Z
Compression	Boundary coding of a selected γ
Physics-analogy	Interpretation of (Γ_adm, A) as open scattering
Training	Parameter tuning so data histories sit low in A
Search (MCTS, ...)	Partial enumeration of Γ_adm
Field dynamics	C evolving through the history, feeding back into A

Why this is the correct level

The smallest object that contains codec, scattering sum, MDL ladder, swarm models, open-kernel port-swap, hyperbolic chart, and Casimir spectrum as projections is the history space h. It is the only thing that is not a chart on something else. Every quantitative improvement comes from computing the sums over Γ_adm more cleanly. The natural unit of analysis is the boundary fibre: a compression ratio is the action of one γ normalised by the log-size of one fibre; an entropy rate is the log-growth of fibres.

State, field, swarm, charge, code, and action are all coordinate systems on one history object.

Full paper (HTML) PDF