Cited by 775Topological Materials and PhenomenaAdvanced Condensed Matter Physics2D Materials and Applications

Hexagonal Warping Effects in the Surface States of the Topological InsulatorBi2Te3

Open access
Fu, Liang·2009·Physical Review Letters

Abstract

A single two-dimensional Dirac fermion state has been recently observed on the surface of the topological insulator Bi2Te3 by angle-resolved photoemission spectroscopy. We study the surface band structure using k x p theory and find an unconventional hexagonal warping term, which is the counterpart of cubic Dresselhaus spin-orbit coupling in rhombohedral structures. We show that this hexagonal warping term naturally explains the observed hexagonal snowflake Fermi surface. The strength of hexagonal warping is characterized by a single parameter, which is extracted from the size of the Fermi surface. We predict a number of testable signatures of hexagonal warping in spectroscopy experiments on Bi2Te3. We also explore the possibility of a spin-density wave due to strong nesting of the Fermi surface.

AI snapshot

Written from the paper’s full text.

TL;DR

This paper derives a hexagonal warping term in the k·p surface-state Hamiltonian of Bi₂Te₃ and shows it explains the observed snowflake-like Fermi surface and its energy-dependent evolution. It extracts the warping strength from ARPES data and predicts several spectroscopic signatures, including out-of-plane spin polarization, magnetic-field-induced gap opening, Friedel-oscillation changes, and a possible spin-density wave.

Key findings

  • An unconventional hexagonal warping term \(H_w\) appears in the surface band structure of Bi₂Te₃; it is the counterpart of cubic Dresselhaus spin-orbit coupling in rhombohedral structures and is solely responsible for hexagonal distortion of the otherwise circular Fermi surface.
  • The warping term reproduces the observed Fermi-surface evolution from snowflake at \(E=0.25\) eV to hexagon and then circle near the Dirac point; fitting the theoretical Fermi momentum along \(\Gamma M\) (\(1.2/a\)) to the experimental value \(0.11\ \text{Å}^{-1}\) gives \(a=10.9\ \text{Å}\), and using \(v=2.55\ \text{eV}\cdot\text{Å}\) gives \(\lambda=250\ \text{eV}\cdot\text{Å}^3\), \(E^*=0.23\ \text{eV}\), and \(E_F=1.2E^*=0.28\ \text{eV}\), agreeing fairly well with the measured \(0.25\ \text{eV}\).
  • Predicted signatures include momentum-dependent out-of-plane spin polarization \(s_z\), reaching as high as 60% of full polarization along \(\Gamma K\) for the Fermi surface in Fig. 1; an in-plane magnetic field shifts the Dirac point and opens a gap \(M\sigma_z=(g_\parallel B_\parallel)^3\sin(3\phi)\sigma_z/E^{*2}\), enabling a quantum Hall effect without Landau levels if \(E_F\) lies in the gap.
  • For a nonmagnetic point defect, leading-order Friedel oscillations are absent for convex constant-energy contours below \(E_c\approx0.69E^*\) but restored above \(E_c\) for non-convex contours; the paper also explores a possible SDW due to strong nesting for \(0.55E^*<E<0.9E^*\), with Landau-Ginzburg analysis giving stripe order for \(u<0\) and a two-dimensional lattice for \(u>0\).

Methods

  • Uses \(k\cdot p\) theory near the \(\Gamma\) point, enforcing \(C_{3v}\) crystal symmetry and time-reversal symmetry to derive a \(2\times2\) effective Hamiltonian up to third order in momentum, including the hexagonal warping term.
  • Compares theoretical constant-energy contours to ARPES data on 0.67% Sn-doped Bi₂Te₃, using measured Fermi velocity \(v=2.55\ \text{eV}\cdot\text{Å}\) and measured Fermi energy \(0.25\ \text{eV}\) to fit the warping parameter.
  • Uses Landau-Ginzburg theory to classify possible spin-density-wave orders from Fermi-surface nesting at wavevectors \(Q_i=2k_F e_i\).

Limitations

  • Stated: the Hamiltonian is said to describe the surface band structure “quite well in a wide energy window at least up to 0.25 eV”; for simplicity the authors discard \(E_0\) and the quadratic velocity correction because they do not significantly change the Fermi-surface shape.
  • Stated: the spin-polarization pattern, magnetic-field gap, and other signatures are predictions “to be tested” in future spin-resolved ARPES or other experiments; the SDW phase is explored theoretically, not observed.
  • Evident from the text: the analysis is a perturbative \(k\cdot p\) expansion truncated at third order, parameter extraction relies on fitting ARPES data, and the note added says the STM test involved a line defect rather than the point defect considered in the paper.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1103/physrevlett.103.266801

Cites

6 of 6 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Claims the paper makes

0 of 7 cited sources checked · 0 supported · 0 flagged

  • Findingown data
    § Hexagonal Warping Effects in the Surface States of Topological Insulator Bi 2 Te 3 · p. 2

    The Fermi surface at EF = 1.2E* is almost identical to the real Fermi surface in 0.67% Sn-doped Bi2Te3 measured in ARPES.

    “Comparing the set of Fermi surfaces in Fig.2a with the real Fermi surface in 0 . 67% Sn-doped Bi 2 Te 3 (Fig.1), we find the Fermi surface at E F = 1 . 2 E ∗ is almost identical”
    Quote in the paperSubstantial claim
  • Findingreasoning
    § Hexagonal Warping Effects in the Surface States of Topological Insulator Bi 2 Te 3 · p. 2

    Using the measured Fermi velocity v = 2.55 eV·Å, the magnitude of the hexagonal warping term is λ = 250 eV·Å^3.

    “Using the measured Fermi velocity v = 2 . 55 eV · ˚ A , we obtain the magnitude of the hexagonal warping term: λ = 250 eV · ˚ A 3 .”
    Quote in the paperSubstantial claim
  • Findingnone given
    § Hexagonal Warping Effects in the Surface States of Topological Insulator Bi 2 Te 3 · p. 3

    The Fermi surface is nearly a hexagon for 0.55E* < E < 0.9E*.

    “We note that the Fermi surface is nearly a hexagon for 0 . 55 E ∗ &lt; E &lt; 0 . 9 E ∗ .”
    Quote in the paperSubstantial claim
  • Findingown data
    § Hexagonal Warping Effects in the Surface States of Topological Insulator Bi 2 Te 3 · p. 2

    Fitting the theoretical Fermi momentum along ΓM (1.2/a) to the experimental value 0.11 Å^-1 gives a = 10.9 Å.

    “By fitting the theoretical value of Fermi momentum along Γ M (1 . 2 /a ) to the experimental one (0 . 11 ˚ A -1 ), we find a = 10 . 9 ˚ A .”
    Quote in the paperSubstantial claim
  • Interpretationnone given
    § Hexagonal Warping Effects in the Surface States of Topological Insulator Bi 2 Te 3 · p. 2

    The form of H(k) is highly constrained by crystal and time reversal symmetry.

    “The form of H ( ⃗ k ) is highly constrained by crystal and time reversal symmetry.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: Bi2Te3 has a rhombohedral crystal structure with space group R-3m.
  • May combine two claims. 18
  • Classified as probably not a substantial claim. 2
  • May combine two claims.

    Findingreasoning
    § Hexagonal Warping Effects in the Surface States of Topological Insulator Bi 2 Te 3 · p. 2

    From λ and v, E* = 0.23 eV and EF = 1.2E* = 0.28 eV, which agrees fairly well with the measured Fermi energy 0.25 eV.

    “From that we find E ∗ = 0 . 23 eV , and E F = 1 . 2 E ∗ = 0 . 28 eV which agrees fairly well with the measured Fermi energy 0 . 25 eV (shown in Fig.1.ii).”
    Quote in the paperSubstantial claim
  • May combine two claims.

    Findingreasoning
    § Hexagonal Warping Effects in the Surface States of Topological Insulator Bi 2 Te 3 · p. 3

    The out-of-plane spin polarization sz is momentum-dependent, given by sz = cos(3θ)/sqrt(cos^2(3θ)+1/(ka)^4), and can reach as high as 60% of the full polarization along ΓK for the Fermi surface in Fig.1.

    “s z is then calculated from (3): s z = cos(3 θ ) / √ cos 2 (3 θ ) + 1 / ( ka ) 4 . The outof-plane spin polarization is momentum-dependent and can reach as high as 60% of the full polarization along Γ K for the Fermi surface in Fig.1.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The pseudo-spin σ_i is proportional to the electron's spin: <sz> ∝ <σz> and <sx,y> ∝ <σx,y>.
  • Classified as probably not a substantial claim.

    Premisecitation
    § Hexagonal Warping Effects in the Surface States of Topological Insulator Bi 2 Te 3 · p. 1

    A single two-dimensional Dirac fermion state has been observed on the surface of the topological insulator Bi2Te3 by angle-resolved photoemission spectroscopy (ARPES).

    “A single two-dimensinoal Dirac fermion state has been recently observed on the surface of topological insulator Bi 2 Te 3 by angle-resolved photoemission spectroscopy (ARPES).”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites [4] Y. L. Chen et al. , Science 325 , 178 (2009) source

      Not checked in the free run.

    • cites [5] D. Hsieh et al. , Nature 460 , 1101 (2009) source

      Not checked in the free run.

  • May combine two claims.

    Interpretationreasoning
    § Hexagonal Warping Effects in the Surface States of Topological Insulator Bi 2 Te 3 · p. 4

    Higher order terms in F break the U(3) symmetry and pick out a particular spatial ordering pattern.

    “We now show that higher order terms in F break the U (3) symmetry and picks out a particular spatial ordering pattern.”
    Quote in the paperSubstantial claim
  • May combine two claims.

    Interpretationreasoning
    § Hexagonal Warping Effects in the Surface States of Topological Insulator Bi 2 Te 3 · p. 3

    For a non-convex constant energy contour above Ec, multiple pairs of stationary points exist, and inter-pair scattering is allowed, so the LDOS oscillation is restored at leading order.

    “Now consider a non-convex constant energy contour above E c . As shown in Fig.3b, multiple pairs of stationary points exist. Since inter-pair scattering is still allowed, the LDOS oscillation will be restored at leading order.”
    Quote in the paperSubstantial claim
  • May combine two claims.

    Interpretationreasoning
    § Hexagonal Warping Effects in the Surface States of Topological Insulator Bi 2 Te 3 · p. 2

    The k·p Hamiltonian H(k) up to third order in k includes a particle-hole asymmetry term E0(k)=k^2/(2m*), a second-order correction to the Dirac velocity v_k=v(1+αk^2), and a hexagonal warping term H_w.

    “We then find that H ( ⃗ k ) must take the following form up to third order in ⃗ k :”
    • Also stated in: § Hexagonal Warping Effects in the Surface States of Topological Insulator Bi 2 Te 3 · p. 2 · plain
      “where E 0 ( k ) = k 2 / (2 m ∗ ) generates particle-hole asymmetry and the Dirac velocity v k = v (1 + αk 2 ) contains a second-order correction. The last term in (3), which we call H w , is most important.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: Bi2Te3 has a rhombohedral crystal structure with space group R-3m.
  • May combine two claims.

    Interpretationreasoning
    § Hexagonal Warping Effects in the Surface States of Topological Insulator Bi 2 Te 3 · p. 4

    The free energy to second order has an emerging U(3) symmetry, so single- and multiple-Q SDWs are degenerate.

    “The free energy (8) to second order has an emerging U (3) symmetry φ iµ → U ij φ jµ . So single- and multipleQ SDWs are degenerate.”
    Quote in the paperSubstantial claim
  • May combine two claims.

    Premisecitation
    § Hexagonal Warping Effects in the Surface States of Topological Insulator Bi 2 Te 3 · p. 4

    Alpichshev et al. imaged with STM the standing wave of surface states on Bi2Te3 near a line defect; the LDOS oscillation was found to exist in the energy range with snow-flake like constant energy contour, but strongly suppressed in the range with circular constant energy contour.

    “During the final stage of this work, we learned that Alpichshev et al.[19] imaged with STM the standing wave of surface states on Bi 2 Te 3 near a line defect (instead of a point defect considered in this work). The LDOS oscillation was found to exist in the energy range with snow-flake like constant energy contour, but strongly suppressed in the range with circular constant energy contour.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [19]

      Not checked in the free run.

  • May combine two claims.

    Interpretationreasoning
    § Hexagonal Warping Effects in the Surface States of Topological Insulator Bi 2 Te 3 · p. 3

    For a convex constant energy contour below Ec, only a single pair of stationary points exist at k and -k; because they carry opposite spins, scattering between them is forbidden by time reversal symmetry, so the LDOS oscillation vanishes at leading order.

    “For a convex constant energy contour below E c as shown in Fig.3a, only a single pair of stationary points exist at ⃗ k and -⃗ k . However, since the two states at ⃗ k and -⃗ k here carry opposite spins, scattering between them is forbidden by time reversal symmetry-a fundamental property of surface states on a topological insulator. The LDOS oscillation then vanishes at leading order.”
    Quote in the paperSubstantial claim
  • May combine two claims.

    Interpretationcitation
    § Hexagonal Warping Effects in the Surface States of Topological Insulator Bi 2 Te 3 · p. 2

    The leading order correction to the surface band dispersion near Γ due to H_w starts at fifth order in k and is angle-dependent.

    “(4) predicts that the leading order correction due to H w starts at fi fth order in k and is angle-dependent:”
    Quote in the paperSubstantial claim
  • May combine two claims.

    Interpretationreasoning
    § Hexagonal Warping Effects in the Surface States of Topological Insulator Bi 2 Te 3 · p. 3

    Since the surface states at k and -k have opposite spins, a charge-density wave cannot connect them and is thus disfavored.

    “Since the surface states at ⃗ k and -⃗ k have opposite spins, a charge-density wave cannot connect them and is thus disfavored.”
    Quote in the paperSubstantial claim
  • May combine two claims.

    Interpretationcitation
    § Hexagonal Warping Effects in the Surface States of Topological Insulator Bi 2 Te 3 · p. 3

    Since spin polarization along ΓM has been found to be almost 100% polarized in ARPES, the doublet at Γ are almost pure spin eigenstates, i.e., sz ≈ <σz>.

    “Since spin polarization along Γ M has been found to be almost 100% polarized in a very recent ARPES experiment[5], we conclude that the the doublet | ψ ↑ , ↓ 〉 at Γ are almost pure spin eigenstates, i.e., s z ≈ 〈 σ z 〉 , which agrees with a theoretical band structure calculation[16].”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites [5] D. Hsieh et al. , Nature 460 , 1101 (2009) source

      Not checked in the free run.

    • cites [16] H. Zhang et al. , Nat. Phys. 5 , 438 (2009) source

      Not checked in the free run.

    • Rests on an assumption: The pseudo-spin σ_i is proportional to the electron's spin: <sz> ∝ <σz> and <sx,y> ∝ <σx,y>.
  • May combine two claims.

    Interpretationreasoning
    § Hexagonal Warping Effects in the Surface States of Topological Insulator Bi 2 Te 3 · p. 4

    For u > 0, |ξ1| = |ξ2| = |ξ3| in the ordered phase, so the SDW forms a two-dimensional lattice.

    “For u &gt; 0, | ξ 1 | = | ξ 2 | = | ξ 3 | in the ordered phase, so that SDW forms a twodimensional lattice.”
    Quote in the paperSubstantial claim
  • May combine two claims.

    Interpretationreasoning
    § Hexagonal Warping Effects in the Surface States of Topological Insulator Bi 2 Te 3 · p. 2

    Although the Hamiltonian H is threefold invariant, the band structure is six-fold symmetric under θ → θ + 2π/6 because of time reversal symmetry.

    “Although the Hamiltonian H is threefold invariant, the band structure is six-fold symmetric under θ → θ +2 π/ 6 because of time reversal symmetry.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: Bi2Te3 has a rhombohedral crystal structure with space group R-3m.
  • May combine two claims.

    Interpretationreasoning
    § Hexagonal Warping Effects in the Surface States of Topological Insulator Bi 2 Te 3 · p. 2

    H_w vanishes in the mirror-symmetric direction ΓM because σz is odd under mirror, and reaches maximum along ΓK.

    “We note that H w ( ⃗ k ) vanishes in mirror-symmetric direction Γ M , because σ z is odd under mirror, and H w ( ⃗ k ) reaches maximum along Γ K .”
    Quote in the paperSubstantial claim
    • Rests on an assumption: Bi2Te3 has a rhombohedral crystal structure with space group R-3m.
  • May combine two claims.

    Interpretationreasoning
    § Hexagonal Warping Effects in the Surface States of Topological Insulator Bi 2 Te 3 · p. 3

    In an in-plane magnetic field B||, the Dirac point is shifted away from Γ to k* = g|| z×B|| /v, and a mass term Mσz = (g||B||)^3 sin(3φ)σz/E*^2 is generated at k*, opening an energy gap.

    “From (3) we find the Dirac point is shifted away from Γ to ⃗ k ∗ ≡ g ‖ ˆ z × ⃗ B ‖ /v . In addition to that, a mass term is generated at ⃗ k ∗ : Mσ z = ( g ‖ B ‖ ) 3 sin(3 ϕ ) σ z /E ∗ 2 ( ϕ is the angle between ⃗ B ‖ and Γ K ), which opens up an energy gap.”
    Quote in the paperSubstantial claim
  • May combine two claims.

    Interpretationreasoning
    § Hexagonal Warping Effects in the Surface States of Topological Insulator Bi 2 Te 3 · p. 2

    The hexagonal warping term H_w is only invariant under three-fold rotation and is solely responsible for the hexagonal distortion of the otherwise circular Fermi surface.

    “Unlike the other terms, H w is only invariant under three-fold rotation (as the Bi 2 Te 3 crystal structure does) and therefore is solely responsible for the hexagonal distortion of the otherwise circular Fermi surface.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: Bi2Te3 has a rhombohedral crystal structure with space group R-3m.
  • May combine two claims.

    Interpretationreasoning
    § Hexagonal Warping Effects in the Surface States of Topological Insulator Bi 2 Te 3 · p. 4

    At fourth order, the anisotropy term F4 = u sum |ξ_i|^4; for u < 0, only one φ_i is nonzero and the SDW forms a one-dimensional stripe that breaks C3 but is invariant under mirror symmetry.

    “At fourth order, we find an anisotropy term F 4 = u ∑ 3 i =1 | ξ i | 4 . The sign of u determines the relative weight of φ i in the ordered phase. For u &lt; 0, only one of φ i , say ξ 1 , is nonzero. The resulting SDW forms a one-dimensional stripe, which breaks C 3 but is invariant under mirror symmetry.”
    Quote in the paperSubstantial claim
  • May combine two claims.

    Premisereasoning
    § Hexagonal Warping Effects in the Surface States of Topological Insulator Bi 2 Te 3 · p. 1

    Two surface bands are observed to touch at the origin of the surface Brillouin zone Γ, and the degeneracy is protected by time-reversal symmetry with the doublet forming a Kramers pair.

    “Two surface bands are observed to touch at the origin of the surface Brillouin zone Γ. The degeneracy is protected by time-reversal symmetry and the doublet | ψ ↑ , ↓ 〉 form a Kramers pair.”
    Quote in the paperSubstantial claim
  • Classified as probably not a substantial claim.

    Interpretationreasoning
    § Hexagonal Warping Effects in the Surface States of Topological Insulator Bi 2 Te 3 · p. 2

    The hexagonal warping term H_w describes cubic spin-orbit coupling at the surface of rhombohedral crystal systems.

    “The hexagonal warping term H w describes cubic spinorbit coupling at the surface of rhombohedral crystal systems, and to the best of our knowledge, it has not been”
    Quote in the paperSubstantial claim
    • Rests on an assumption: Bi2Te3 has a rhombohedral crystal structure with space group R-3m.
20 of 20
  • The paper relies on this without support.

    Bi2Te3 has a rhombohedral crystal structure with space group R-3m.

    “Bi 2 Te 3 has a rhombohedral crystal structure with space group R 3 ¯ m .”
  • The paper relies on this without support.

    The pseudo-spin σ_i is proportional to the electron's spin: <sz> ∝ <σz> and <sx,y> ∝ <σx,y>.

    “Here the pseudo-spin σ i is proportional to electron's spin: 〈 s z 〉 ∝ 〈 σ z 〉 and 〈 s x,y 〉 ∝ 〈 σ x,y 〉 .”
  • The paper relies on this without support.

    The susceptibility matrix χ is positive definite in the normal state.

    “χ µν is positive definite in the normal state.”
3 of 3

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?

  • Magnetic impurity such as Fe on the surface of Bi2Se3 opens a gap at the Dirac point.

    “Magnetic impurity such as Fe on the surface of Bi 2 Se 3 opens a gap at the Dirac point (Fig.5) (22,23,38).”

    From Three-Dimensional Topological Insulators (2011) · cites this paper as [38]

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

  • Reported hexagonal-warping strength does not match the fitted parameters

    Major

    The stated characteristic length satisfies λ = v a2. With a = 10.9 Å and v = 2.55 eV·Å, this gives λ ≈ 303 eV·Å3, not the reported 250 eV·Å3.

    Reported 250 · recomputed 302.9655

1 of 1

Cites

Flagged first, then by citations

Research works printed but not linked: 11

  1. 10.1126/science.1167733 · DOI · Paper page
  2. 10.1038/nature08234 · DOI · Paper page
  3. 10.1103/PhysRevB.75.121306 · DOI · Paper page
  4. 10.1016/0370-2693(83)91529-0 · DOI · Paper page
  5. 10.1103/PhysRevB.76.045302 · DOI · Paper page
  6. 10.1103/PhysRevB.78.195424 · DOI · Paper page
  7. 10.1063/1.3200237 · DOI
  8. 10.1103/PhysRevB.61.2981 · DOI
  9. 10.1038/nphys1270 · DOI · Paper page
  10. 10.1103/PhysRevLett.61.2015 · DOI · Paper page
  11. 10.1103/PhysRev.149.519 · DOI

Cited by

Sorted by citations

Authors
Fu, L.
Affiliations
Harvard University (US)
Corresponding author
Fu, Liang
Venue
Physical Review Letters
Published
2009-12-21
Volume
103
Issue
26
Pages
266801-266801
ISSN
  • 0031-9007 (ISSN-L)
  • 1079-7114
OpenAlex
W1813652558
Cited by
775
References
6
Citations by year
2026: 38 · 2025: 41 · 2024: 31 · 2023: 46 · 2022: 38 · 2021: 57 · 2020: 38 · 2019: 42 · 2018: 44 · 2017: 41 · 2016: 45 · 2015: 49 · 2014: 54 · 2013: 71 · 2012: 66
FWCI
25.92
Type
journal article
Language
en
Open access
Locations
  • Physical Review Letters (publishedVersion)
  • arXiv (Cornell University) (publishedVersion)
  • PubMed (publishedVersion)
  • arXiv (Cornell University)
Topics
  • Physical Sciences › Physics and Astronomy › Atomic and Molecular Physics, and Optics › Topological Materials and Phenomena
  • Physical Sciences › Physics and Astronomy › Condensed Matter Physics › Advanced Condensed Matter Physics
  • Physical Sciences › Materials Science › Materials Chemistry › 2D Materials and Applications
Keywords
Bi2Te3, topological insulators, hexagonal warping, surface states, two-dimensional Dirac fermions, k·p theory, Dresselhaus spin-orbit coupling, Fermi surface, spin density wave, Fermi surface nesting, rhombohedral structure

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