Cited by 77Noncommutative and Quantum Gravity TheoriesRelativity and Gravitational TheoryQuantum Mechanics and Applications

Almost ideal clocks in quantum cosmology: a brief derivation of time

Open access
Marolf, Donald·1995·Classical and Quantum Gravity

Abstract

A formalism for quantizing time reparametrization invariant dynamics is considered and applied to systems which contain an `almost ideal clock'. Previously, this formalism was successfully applied to the Bianchi models and, while it contains no fundamental notion of `time' or `evolution', the approach does contain a notion of correlations. Using correlations with the almost ideal clock to introduce a notion of time, we derive the complete formalism of external time quantum mechanics. The explicit relation between the physical Hilbert space and the usual Hilbert space for a system with an external time is also derived and the limit of an ideal clock is found to be closely associated with the Klein - Gordon inner product and the Newton - Wigner formalism.

Integrity signals

Registry

Crossref · 10.1088/0264-9381/12/10/007

Cites

24 of 24 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Claims the paper makes

0 of 24 cited sources checked · 0 supported · 0 flagged

  • Findingnone given

    The factorization H0 = Hp ⊗ Hc is not independent of the parameter time t.

    “In general, the decomposition H 0 = H p ⊗ t H c will depend on a choice of parameter time t.”
    Quote in the paperNot a substantial claim
  • Findingreasoning

    Except near a turning point, q_c(t) is roughly q_c(0)+n, with n the integer count of left-turning-point visits by q since t=0.

    “Thus, unless q is near a turning point, q c (t) ∼ = q c (0)+n where n is the (integer) number of times that q has reached its left turning point since t = 0.”
    Quote in the paperSubstantial claim
    • The authors state this as an assumption: The ambiguity of the pendulum as a clock is removed by coupling it to a counter that increments each complete oscillation.
    • The authors state this as an assumption: V(q) has no more than a single critical point, and if it exists it is a global minimum.
  • Findingreasoning

    To leading order in p_c, the A±τ,n operators are symmetric on H_aux.

    “Thus, the A ± τ,n are symmetric on H aux to lowest order in p c .”
    Quote in the paperSubstantial claim
  • Findingreasoning

    The expectation value of Aτ in a physical state equals 2πħ times an auxiliary inner product involving θ(P)√P χ0 τ A √P θ(P).

    “ψ|A τ |ψ phys = 2πh ψ|θ(P ) √ P χ 0 τ A √ P θ (P )|ψ aux (4.5)”
    Quote in the paperSubstantial claim
  • Findingreasoning

    When A is constructed from non-clock variables, A_{τ+δ} is e^{iδPτ} Aτ e^{-iδPτ} multiplied by (1+O(ε)+O(h)).

    “A τ +δ = e iδP τ A τ e -iδP τ (1 + O ( ) + O (h))”
    Quote in the paperSubstantial claim
    • The authors state this as an assumption: The clock coordinate is far from classical turning points so that classical q(t) is single-valued.
    • The authors state this as an assumption: |p_c| is small in the sense that only states with spectral support in P_c concentrated where p_c << E are considered, E being the splitting between energy levels of H_p.
  • May combine two claims. 26
  • Classified as probably not a substantial claim. 5
  • The quote only partly states this claim. 2
  • May combine two claims.

    Findingreasoning

    The normalization of |p_c is ε-dependent and becomes divergent at ε=0.

    “Note that the normalization of |p c is -dependent and that, in fact, (3.8) diverges at = 0.”
    Quote in the paperNot a substantial claim
  • May combine two claims.

    Findingreasoning

    At p_c=p_c the matrix element p_c|χ_a|p_c is 1, and it decays with width 2/h.

    “which is 1 at p c = p c and decays away from p c = p c with a width of 2/h .”
    Quote in the paperSubstantial claim
  • May combine two claims.

    Findingnone given

    The commutator of Q_c with ∂Q(t)/∂t has order p_c and is negligible.

    “Note that the commutator of Q c with ∂Q(t)/∂t is of order p c and so negligible.”
    Quote in the paperNot a substantial claim
  • May combine two claims.

    Findingreasoning

    Classically, the counter coordinate moves by ε each pendulum swing and remains nearly constant when |q-q0(E)| is much larger than λ.

    “in the classical version of this system, (3.3) is constructed so that the counter coordinate q c advances a distance with each oscillation of the pendulum but is essentially constant when |qq 0 (E)| λ.”
    Quote in the paperSubstantial claim
    • The authors state this as an assumption: V(q) has no more than a single critical point, and if it exists it is a global minimum.
  • May combine two claims.

    Findingreasoning

    For conserved clock momentum, A_{τ+δ}=e^{iδP} Aτ e^{-iδP}, yielding exact unitary evolution.

    “Using this property, a brief study of the observables A τ defined by (3.4) shows that A τ +δ = e iδP A τ e -iδP and we have exact unitary evolution.”
    Quote in the paperSubstantial claim
  • May combine two claims.

    Findingcitation

    When V(Q)=0, the WKB approximation is exact and P θ(P)|ψ = sqrt(-H1) θ(P)|ψ.

    “However, when V (Q) = 0 the WKB approximation is exact and P θ (P )|ψ = √ -H 1 θ(P )|ψ .”
    Quote in the paperSubstantial claim
  • May combine two claims.

    Findingreasoning

    When [P,H]=0, the clock momentum is conserved, P(t)=P is constant, and it defines a physical operator.

    “This occurs when the clock momentum P is conserved ([P , H ] = 0) so that P (t) = P is independent of t and defines a physical operator.”
    Quote in the paperNot a substantial claim
  • May combine two claims.

    Findingreasoning

    To this approximation the existence operator 1lτ equals the identity, and therefore ∂1lτ/∂τ=0.

    “Thus, the 'existence' operator 1l τ is the identity to this approximation and, in particular, ∂1l τ /∂τ = 0.”
    Quote in the paperSubstantial claim
    • The authors state this as an assumption: |p_c| is small in the sense that only states with spectral support in P_c concentrated where p_c << E are considered, E being the splitting between energy levels of H_p.
  • May combine two claims.

    Findingcitation

    When V(Q)=0 and H1=g_ij P⊥i P⊥j with a negative definite constant metric, (Q_i⊥)+τ and (P⊥_j)+τ are the usual positive-frequency Newton-Wigner operators and satisfy the canonical commutation relations exactly.

    “While it does not follow from the above argument, the results of section 4.2 of [1] show that when V (Q) = 0 and H 1 = ij g ij P ⊥ i P ⊥ h (for any negative definite constant metric g ij ) the (Q i ⊥ ) + τ and the (P ⊥ j ) + τ are just the usual positive frequency Newton-Wigner operators and so satisfy the canonical commutation relations exactly.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [1] Marolf D, 1995, Class. Quantum Grav. source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingreasoning

    It is possible that ih∂1lτ/∂τ = [1lτ,P] is nonzero.

    “Note also that ih∂1l τ /∂τ = [1l τ , P ] may not vanish †.”
    Quote in the paperNot a substantial claim
  • May combine two claims. The quote only partly states this claim.

    Findingreasoning

    From (2.6) and p_c small, P^mτ acting on |n,p_c,-E(n,p_c),a gives (2E(n,p_c)-2V(τ))^{m/2} times that state to the working accuracy, exactly when V=0.

    “it follows from (2.6) and the assumption that p c 1 that P m τ |n, p c , -E(n, p c ), a = (2E(n, p c ) -2V (τ )) m/2 |n, p c , -E(n, p c ), a (4.2)”
    Quote in the paperSubstantial claim
    • The authors state this as an assumption: |p_c| is small in the sense that only states with spectral support in P_c concentrated where p_c << E are considered, E being the splitting between energy levels of H_p.
  • May combine two claims.

    Findingreasoning

    ρ(τ)=|ψ;τ><ψ;τ| with |ψ;τ=σ_τ θ(P)√P η|ψ, yielding ψ|Aτ|ψ = <ψ;τ|A|ψ;τ>.

    “We may then write ρ(τ ) = |ψ; τ ψ; τ | where |ψ; τ = σ τ θ(P ) √ P η|ψ (4.9) so that we have derived ψ|A τ |ψ = ψ; τ|A|ψ; τ .”
    Quote in the paperSubstantial claim
  • May combine two claims.

    Findingreasoning

    Aτ Bτ equals (AB)τ to leading order in ħ, so the full algebra of external time quantum mechanics is reproduced to this order.

    “Finally, we note that the arguments applied to the type I terms above can also be used to show that A τ B τ = (AB) τ to leading order in h so that, to this order, the full algebra associated with external time quantum mechanics has been reproduced.”
    Quote in the paperSubstantial claim
  • May combine two claims.

    Findingreasoning

    On the physical Hilbert space, the commutator of Pτ with Aτ equals -iħ times the τ derivative of Aτ, which verifies the evolution law (4.1).

    “Thus, we have [P τ , A τ ] = -ih∂A τ /∂τ on H phys , verifying the evolution law (4.1).”
    Quote in the paperSubstantial claim
  • May combine two claims.

    Findingreasoning

    When [A,H1]=0, φ|θ(P)A[P,τ]θ(P)|ψ=0, so (4.10) is exact.

    “Thus, when [A, H 1 ] = 0, we have φ|θ(P )A[P , τ ]θ(P )|ψ = 0 so that (4.10) is exact.”
    Quote in the paperNot a substantial claim
  • May combine two claims.

    Interpretationreasoning

    When V(q) confines q to a finite region, H_p is a pendulum, not a clock; then q specifies several possible moments, not a single moment.

    “if V (q) confines q to some finite region, what H p describes is not a clock but a pendulum; that is, a part of a clock, or a clock that measures time only modulo some (possibly state dependent) period.Thus, specifying the coordinate q determines not a unique moment of time but a discrete set of such moments.”
    Quote in the paperSubstantial claim
  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Interpretationnone given

    Classically, a pendulum of this kind can serve as a clock only over a chosen half-cycle.

    “Even classically, such a pendulum can only be used as a clock during some agreed upon half of an oscillation”
    Quote in the paperNot a substantial claim
  • Classified as probably not a substantial claim.

    Premisenone given

    Operators on H_phys as defined in (2.6) are independent of which t is selected.

    “Since, however, the definition (2.6) of the operators on H phys is independent of the choice of t, no harm is done by choosing t = 0.”
    Quote in the paperSubstantial claim
    • The authors state this as an assumption: The auxiliary Hilbert space has a basis for a spectral interval [a,b] of H containing 0, and H has continuous spectrum with no point spectrum in that interval.
  • May combine two claims.

    Premisecitation

    The almost local formalism supplies a positive-definite inner product on a physical Hilbert space and a complete set of symmetric quantum observables (perennials) that capture the value of A at the time Q=τ.

    “it provides [1,22] a positive-definite inner product on a physical Hilbert space H phys as well as a complete set of well-defined symmetric quantum observables on this space (perennials in the language of [18]) that, in a certain sense, capture the notion of 'the value of quantity A at the point in time at which quantity Q takes a specified value τ '.”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites [1] Marolf D, 1995, Class. Quantum Grav. source

      Not checked in the free run.

    • cites [22] Ryan M, 1975

      Not checked in the free run.

  • The quote only partly states this claim.

    Study design and conductnone given

    Approximation (3.5) is expected to be valid provided V(τ) is much smaller than E(n,p_c).

    “We expect this approximation to be valid when V (τ ) E(n, p c ) ≡ (n; p c |H p c 0 |n; p c ).”
    Quote in the paperNot a substantial claim
    • The authors state this as an assumption: The clock coordinate is far from classical turning points so that classical q(t) is single-valued.
  • May combine two claims.

    Interpretationreasoning

    The lapse N can differ from the metric lapse and may be rescaled nontrivially.

    “N may not correspond exactly to the metric-defined lapse but may be rescaled in some nontrivial way.”
    Quote in the paperSubstantial claim
  • May combine two claims.

    Interpretationreasoning

    For a homogeneous comoving fleet of ideal metric clocks in a homogeneous cosmology, unitarity may hold exactly while conservation of existence generally fails.

    “We again recall (see the appendix) that for a homogeneous comoving fleet of ideal metric clocks in a homogeneous cosmology, unitarity may hold exactly but we will not, in general, have conservation of existence.”
    Quote in the paperSubstantial claim
    • The authors state this as an assumption: The clocks in the fleet are evenly distributed with respect to the metric on each spatial slice.
    • The authors state this as an assumption: The clock fleet is comoving, with its 4-velocity proportional to the normal to the family of slices.
  • May combine two claims.

    Interpretationreasoning

    In the ideal clock limit, Aτ become positive-frequency Newton-Wigner operators, and the approach agrees with Wald [32].

    “Similarly, the operators A τ become the positive frequency Newton-Wigner operators in the ideal clock limit and, in this limit, our approach will agree with that of Wald [32].”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [32] Liboff R, 1980

      Not checked in the free run.

  • May combine two claims.

    Interpretationreasoning

    If the square-root argument turns negative, the approximations break down and the pendulum should not be expected to serve well as a clock.

    “In this regime, the approximations considered are not valid and, from general arguments, we have no reason to expect that our pendulum should function well as a clock.”
    Quote in the paperSubstantial claim
  • May combine two claims.

    Interpretationreasoning

    The adiabatic approximation should hold when condition (5.3) holds, which is expected when e^{3α} is large.

    “Thus, the adiabatic approximation should hold when Ṗ / α P .Using h Ȧ ∝ [A, NH ] and the fact that H is typically of a form (see [1,22]) such that [α, H ] = hP α , this condition is V (Q) P α e 3α P (5.3) which we may expect to hold when e 3α is large.”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites [1] Marolf D, 1995, Class. Quantum Grav. source

      Not checked in the free run.

    • cites [22] Ryan M, 1975

      Not checked in the free run.

  • May combine two claims.

    Interpretationreasoning

    Real clocks are used only insofar as they measure proper time and agree, i.e., to the extent that dispersion can be neglected.

    “In practice, real clocks are only used to the extent that they measure proper time and thus agree with each other; that is, to the extent that the dispersion above may be neglected.”
    Quote in the paperSubstantial claim
  • Classified as probably not a substantial claim.

    Study design and conductreasoning

    The entire derivation used only the approximations of section 3, namely the ideal limit of an almost ideal clock from section 3.1.

    “Recall that the entire derivation followed using only the approximations of section 3 that we have an ideal limit of an almost ideal clock as described in 3.1.”
    Quote in the paperSubstantial claim
  • May combine two claims.

    Interpretationreasoning

    Unlike standard WKB derivations, our assumptions do not require the clock to be large or massive, so non-clock degrees can be more than a small perturbation and the clock need not have nearly zero energy.

    “In particular, in contrast to standard derivations [8][9][10][11][12][13][14] of evolution through the use of the WKB approximation, our assumptions do not imply that the clock is large or massive so that the non-clock degrees of freedom may, in this approach, represent more than a small perturbation and the clock need not be in a state with nearly zero energy †.”
    Quote in the paperSubstantial claim0 of 3 sources checked
    • cites [8] Lapchinsky V G, 1979, Acta. Phys. Pol.

      Not checked in the free run.

    • cites [9] source

      Not checked in the free run.

    • cites [10] Banks T, 1985, Nucl. Phys. source

      Not checked in the free run.

  • May combine two claims.

    Interpretationreasoning

    At ε=0 the spectrum of H0 is continuous, and WKB states normalized by (3.9) obey E|E=δ(E-E); hence counter-free results apply at ε=0.

    “In this case, the spectrum of H 0 is already continuous and, while the label n is now inappropriate, the WKB states normalized as in (3.9) satisfy E|E = δ(E -E ), so that all results below which do not explicitly involve the counter may also be used in this case with = 0.”
    Quote in the paperNot a substantial claim
  • May combine two claims.

    Interpretationreasoning

    P and Pτ = P 1lτ differ unless 1lτ=1l, and that equality fails in interesting systems.

    “However, P and P τ = P 1l τ (where 1l τ is also defined by (3.4)) are different operators unless we have 1l τ = 1l which, as the appendix describes, is not the case for interesting such systems.”
    Quote in the paperNot a substantial claim
  • Classified as probably not a substantial claim. May combine two claims.

    Interpretationcitation

    The complete formalism of [1] allowed addressing unitary evolution in the physical inner product and deriving the complete map from the physical Hilbert space to the external time quantum mechanics Hilbert space.

    “In addition, the complete formalism of [1] has allowed a more extensive treatment than was previously possible-we were able to address the issue of unitary evolution in the physical inner product and to derive the complete map from the physical Hilbert space to the corresponding Hilbert space of external time quantum mechanics ‡.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [1] Marolf D, 1995, Class. Quantum Grav. source

      Not checked in the free run.

31 of 31
  • The authors state this as an assumption.

    The auxiliary Hilbert space has a basis for a spectral interval [a,b] of H containing 0, and H has continuous spectrum with no point spectrum in that interval.

    “Precisely, we assume that the Hilbert space H aux has a basis {|n, E : n ∈ Z, E ∈ [a, b]} for the part of H aux corresponding to some spectral interval [a, b] of H with 0 ∈ [a, b].”
  • The authors state this as an assumption.

    The unphysical decoupled-clock constraint can be approximated by systems differing only by terms treatable adiabatically or perturbatively.

    “While this situation is unphysical, we assume that it may in turn be approximated by systems which differ from (3.2) only by terms that may be treated adiabatically or in perturbation theory.”
  • The authors state this as an assumption.

    The clock coordinate is far from classical turning points so that classical q(t) is single-valued.

    “For such systems, we must work in the approximation that the clock coordinate q is far from the classical turning points so that the classical q(t) is a single-valued function of time.”
  • The authors state this as an assumption.

    The ambiguity of the pendulum as a clock is removed by coupling it to a counter that increments each complete oscillation.

    “Following the model of a grandfather clock, we will assume that this ambiguity is removed by coupling the pendulum to some counter which increments with each complete oscillation.”
  • The authors state this as an assumption.

    V(q) has no more than a single critical point, and if it exists it is a global minimum.

    “where we have assumed that V (q) has no more than a single critical point and that this point (if it exists) is a global minimum.”
  • The authors state this as an assumption.

    |p_c| is small in the sense that only states with spectral support in P_c concentrated where p_c << E are considered, E being the splitting between energy levels of H_p.

    “First, we assume that | p c | is small in the sense that we consider only states |ψ ∈ H phys whose spectral support in P c is concentrated in the region with p c E, where E is the splitting between the energy levels of H p on which the physical 'state' |ψ ∈ H aux is concentrated.”
  • The authors state this as an assumption.

    For a chosen pendulum position τ, for each eigenvalue E of H0 either |τ - q_t| >> h/√E for any solution V(q_t)=E or the corresponding projection of |ψ is negligible.

    “Choosing some pendulum position τ , we assume that for each eigenvalue E of H 0 either |τ -q t | h/ √ E for any solution of V (q t ) = E or that the corresponding projection of |ψ is negligible.”
  • The authors state this as an assumption.

    The width λ in (3.3) is large (λ >> h/√E) so that the turning point need not be defined too precisely.

    “We also assume that the width λ in (3.3) is large (λ h/ √ E) so that the notion of a turning point need not be defined too precisely.”
  • The authors state this as an assumption.

    The projection of |ψ to the subspace where Q=τ has negligible projections onto eigenvalues 0 and of Q_c so the counter may be read clearly.

    “Finally, we ask that the projection of |ψ to the subspace where Q = τ have negligible projections onto the eigenvalues 0 and of Q c so that the counter may be read clearly.”
  • The authors state this as an assumption.

    The clocks in the fleet are evenly distributed with respect to the metric on each spatial slice.

    “such that the clocks are evenly distributed with respect to the metric on each slice.”
  • The authors state this as an assumption.

    The clock fleet is comoving, with its 4-velocity proportional to the normal to the family of slices.

    “Let us also assume that the clocks are comoving in the sense that the 4-velocity field U α of the clock fleet is proportional to the normal field n α of the family of slices t .”
  • The authors state this as an assumption.

    The clock is coupled to the description of the gravitational field used in [1] or [22] when considering metric-defined proper time.

    “Had the clock instead measured metric-defined proper time (assuming that it is coupled to the description of the gravitational field used in [1] or [22]), our system would be described”
12 of 12

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

  • In [15], the authors introduce an algebra of observables for complex Ashtekar gravity and the Maxwell field coupled to fermions, with fermions coupling by insertions at endpoints of Wilson lines, and compute the Poisson and commutator algebra of this overcomplete set of gauge invariant observables.

    “in [15] the authors introduce an algebra of observables for complex Ashtekar gravity and the Maxwell field coupled to fermions. The fermions couple by making insertions at the endpoints of Wilson lines. The authors compute the Poisson and commutator algebra of this (overcomplete) set of gauge invariant observables.”
  • The authors of [21] formulated a kinematical framework for diffeomorphism invariant theories of fermions coupling to real-valued Ashtekar gravity and other gauge fields, defining a natural inner product by the Berezin integral, showing a class of operators are well-defined, and indicating how to define diffeomorphism invariant distributions involving fermions.

    “The authors of [21] have followed this trend and formulated a kinematical framework for diffeomorphism invariant theories of fermions coupling to real-valued Ashtekar gravity and other gauge fields. That is, they define a natural inner product by means of the Berezin integral, show that a certain class of the operators defined in [15] are well-defined on it and indicate how to define diffeomorphis”
  • Difficulties with the awkward reality structure of the fermion field make the framework of [21] incomplete: most interesting operators defined in [15] do not have an adjoint on the Hilbert space defined, and this prevented proving that their inner product is selected by the classical reality conditions.

    “However, certain difficulties having to do with the awkward reality structure of the fermion field make their framework incomplete in the sense that most of the interesting operators defined in [15] do not have an adjoint on the Hilbert space defined. In particular, this prevented the authors from proving that their inner product is selected by the classical reality conditions.”
  • The result that the inner product is uniquely selected by reality conditions can be established at the non-gauge-invariant level, without dealing with complicated path observables defined in [15,21].

    “We do not need to deal with the complicated path observables defined in [15, 21] in order to establish this result because we can prove everything at the non-gauge-invariant level.”

Cites

Flagged first, then by citations

Research works printed but not linked: 2

  1. Lapchinsky V G, 1979, Acta. Phys. Pol., 10, 1041
  2. Jantzen R, 1979, 64, 211

Cited by

Sorted by citations

Authors
Marolf, D.
Affiliations
University of California, Santa Barbara (US)
Corresponding author
Marolf, Donald
Venue
Classical and Quantum Gravity
Published
1995-10-01
Volume
12
Issue
10
Pages
2469-2486
ISSN
  • 0264-9381 (ISSN-L)
  • 1361-6382
OpenAlex
W2131407107
Cited by
77
References
24
Citations by year
2026: 4 · 2025: 4 · 2024: 3 · 2023: 1 · 2022: 2 · 2021: 2 · 2020: 1 · 2019: 5 · 2017: 2 · 2016: 1 · 2015: 3 · 2014: 4 · 2012: 7
FWCI
2.34
Type
journal article
Language
en
Open access
Locations
  • Classical and Quantum Gravity (publishedVersion)
  • arXiv (Cornell University) (submittedVersion)
  • arXiv (Cornell University)
Funders
National Science Foundation (9008502)
Topics
  • Physical Sciences › Physics and Astronomy › Statistical and Nonlinear Physics › Noncommutative and Quantum Gravity Theories
  • Physical Sciences › Physics and Astronomy › Astronomy and Astrophysics › Relativity and Gravitational Theory
  • Physical Sciences › Physics and Astronomy › Atomic and Molecular Physics, and Optics › Quantum Mechanics and Applications
Keywords
quantum cosmology, time reparametrization invariance, physical Hilbert space, Bianchi models, correlation, quantum time

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Claims read by deepseek-v4.1-flash under definition version 3 (extract-v3) on 2026-10-06.

Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-10-06.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review