greatitself.
Continue
research27 min studyLesson 763 in Physics

Connect Euclidean slices to a positive Hilbert space

Distinguish positive weights, positive operators, and assumptions behind reconstruction.

Before you begin
On this page

The question for today

What you’ll understand

  • Relate transfer eigenvalues to energy differences.
  • Explain reflection positivity beyond pointwise positive weights.
  • Use normalized Haar integration without redundant gauge fixing.

A Euclidean lattice integral is useful for quantum physics when its time-slice structure supports a positive-norm Hilbert space and a sensible Hamiltonian. A transfer matrix makes that connection concrete. Reflection positivity is one of the structural conditions behind the construction, and is stronger than requiring the integrand to be nonnegative at each configuration. Compact group integration and gauge constraints also enter the definition. These statements apply to specified actions and parameter domains, not automatically to every improved discretization carrying the label lattice gauge theory.

Translate one time slice

Separate spatial and temporal lattice spacings as aS and aT when necessary. The transfer operator advances by one Euclidean time step. For a positive self-adjoint operator, its eigenvalues can be related to exponentials of real energies. Dividing by the vacuum eigenvalue removes an arbitrary vacuum normalization and gives energy gaps. Gauge-invariant states satisfy the lattice Gauss constraint; redundant local color frames are not additional physical states to count in the spectrum.

For the standard pure-gauge Wilson action with β>0 and suitable reflection-symmetric temporal boundaries, a finite spatial lattice with compact links supports the usual positive transfer construction. Positivity and locality conditions must be rechecked for other actions. Some improved actions can violate simple finite-spacing reflection positivity even if they are designed to approach a useful continuum theory. Short-time artifacts then require particular caution; one cannot infer a universally positive spectral decomposition merely from a positive numerical action value.

T=e−aTH,aT(En−E0)=−log⁡λnλ0,λn>0\mathcal T=e^{-a_T H},\qquad a_T(E_n-E_0)=-\log\frac{\lambda_n}{\lambda_0},\qquad \lambda_n>0
Natural units ℏ=c=1 are used; aT is a length and H has inverse-length units. The formula assumes the indicated positive transfer eigenvalues and an appropriate self-adjoint physical-sector operator.↔ Scroll wide equations horizontally.

Reflect an observable, not just a configuration weight

Reflection positivity tests a quadratic form built from an observable supported on one side of a Euclidean time reflection and its reflected complex conjugate. Nonnegativity allows an inner product after null vectors are identified. This is a statement about the entire correlation structure and reflection map. Ordinary positivity of exp(−S) ensures an integration weight, but alone does not guarantee positivity of the associated transfer operator’s spectrum.

For example, a symmetric two-state matrix can have every entry positive yet possess one negative eigenvalue. As a stochastic matrix it may still define valid probabilities, with an alternating approach to stationarity. As a candidate positive quantum transfer operator it fails the required positivity test. Confusing these two uses of the word positive would turn an oscillatory Markov relaxation mode into an unphysical real-energy logarithm.

⟨(ΘF)F⟩E≥0,F supported at positive Euclidean times\langle(\Theta F)F\rangle_E\ge0,\qquad F\ \text{supported at positive Euclidean times}
Θ includes time reflection and complex conjugation. This schematic condition presupposes specified reflection planes, action, boundary conditions, and gauge-invariant observable algebra; it is not equivalent to S being real.↔ Scroll wide equations horizontally.

Haar measure keeps compact integration finite

For each independent link integrate with normalized Haar measure, invariant under both left and right group multiplication. On a finite lattice the product of compact normalized measures is finite, so gauge-related configurations do not generate an infinite gauge-volume divergence. Gauge-invariant expectations can be defined directly without gauge fixing. A perturbative expansion in unconstrained Lie-algebra variables changes that situation and generally needs its usual gauge-fixing treatment.

The Markov lab contrasts probability transitions with operator positivity. Reset and set p=q=0.25, then Step once; predict probability of state 1 equal to 0.25 and stationary value 0.5. Reset and set p=q=0.8, then Step once; predict 0.8 with the same stationary value. The latter chain has an alternating relaxation mode. These are exact two-state distributions, not lattice transfer matrices, Haar samples, or a reflection-positivity test for gauge fields.

∫GdU=1,d(VUW)=dU,∫dU  UijUkl∗=1Nδikδjlfor fundamental SU(N)\int_GdU=1,\qquad d(VUW)=dU,\qquad \int dU\;U_{ij}U^*_{kl}=\frac1N\delta_{ik}\delta_{jl}\quad\text{for fundamental SU}(N)
V,W are fixed group elements and Haar measure is normalized. The final finite group identity expresses orthogonality of fundamental matrix elements and supplies a useful link-integral normalization check.↔ Scroll wide equations horizontally.

Work it through

A positive two-state spectrum

As a finite illustrative transfer operator, use T=[[0.75,0.25],[0.25,0.75]]. Find its two eigenvalues and the energy gap in units of inverse temporal spacing.

  1. The equal-component vector is an eigenvector with eigenvalue 1 because the row sums are one. The opposite-component vector has eigenvalue 0.75−0.25=0.5.

  2. Both eigenvalues are positive, so this symmetric matrix is a positive operator. Normalize the vacuum eigenvalue to one and identify the excitation ratio as 1/2.

  3. The gap satisfies aTΔE=−ln(1/2)=ln2. This simple spectrum illustrates the logarithm but carries no gauge links or Gauss-law projection of its own.

The result

The eigenvalues are 1 and 1/2, giving ΔE=(ln2)/aT.

The living notebook

Explore the model

Interactive lab

Interactive lab · markov

Contrast entry positivity and spectral positivity

Reset and set p=0.25,q=0.25, then Step once; predict P₁=0.25. Reset and set p=0.8,q=0.8, then Step once; predict P₁=0.8. Both stationary probabilities are 0.5.

Contrast entry positivity and spectral positivityExact finite-step probabilities for a two-state chain starting in state 0, not sampled trajectories. Playback advances two transitions per second. When p = q = 0 every distribution is stationary; when p = q = 1 the chain alternates forever. A single zero transition creates an absorbing state. Current readings: P(state 1) at step 0: 0%; Stationary P(state 1): 60%; Chain behavior: Monotone convergence. All parameters are adjustable with the labeled controls below.Two-state Markov distribution after 0 steps: state zero probability 1, state one probability 0.Initial distribution (1, 0) · step 0p = 0.3q = 0.201100%0%1−p1−qP(0)P(1)Monotone convergence
P(state 0)P(state 1)Stationary distribution
P(state 1) at step 0
0%
Stationary P(state 1)
60%
Chain behavior
Monotone convergence
t = 0 s

Exact finite-step probabilities for a two-state chain starting in state 0, not sampled trajectories. Playback advances two transitions per second. When p = q = 0 every distribution is stationary; when p = q = 1 the chain alternates forever. A single zero transition creates an absorbing state. Parameter changes pause playback and recompute the model at the current elapsed time. Reset restores both parameters and returns time to zero. Playback starts paused. Adjust either parameter, then choose Play or Step.

Look forThe second stochastic chain alternates around equilibrium. These probability matrices do not implement quantum transfer evolution or certify reflection positivity.

Pause & reason

Check your understanding

Which condition is stronger than merely having a nonnegative Euclidean configuration weight?

Your turn

Try it yourself

Consider T=[[0.2,0.8],[0.8,0.2]]. Find its eigenvalues. Explain why it is a valid stochastic matrix but cannot be the positive transfer operator assumed in the displayed logarithm.

Take a moment to make a prediction or work it out on paper. Open a hint when you need one.

Give me a hint

Use the equal and opposite component vectors again; check eigenvalues rather than only entries.

Compare with the solution

The eigenvalues are 1 and −0.6. Its entries are nonnegative and row sums equal one, so it gives valid two-state Markov transitions. The negative eigenvalue violates positive-operator transfer positivity and has no real logarithmic energy in T=exp(−aTH) with self-adjoint H. Positive matrix entries do not settle that question.

Keep the idea

Take it with you

  • Transfer eigenvalue ratios determine physical energy gaps under positivity assumptions.
  • Reflection positivity concerns correlations across a time reflection.
  • Compact normalized Haar integration defines finite link integrals without mandatory gauge fixing.

Read further

Original sources and deeper explanations to follow at your own pace.

Optional progress, saved in this browser when available. Revisit any lesson at any time.

Lesson not marked complete.