Candidate verdicts

TESS Objects of Interest (TOIs) whose TFOPWG disposition is still PC (planet candidate) are run through the full pipeline and the vetting tests (scripts/vet_toi_candidates.py).

Selection rule (deterministic, applied to the TOI table at run time): disposition PC; 1 d < P < 15 d; TESS magnitude ≤ 11; catalog depth ≥ 800 ppm; ranked by the S/N proxy depth × 10^(−0.2 (Tmag − 10)) × √(27.4 d / P), which favors bright, deep, frequently transiting candidates; at most one TOI per star; the first five with SPOC 2-minute light curves filed under their own TIC ID are used. The TOI table was queried on 28 September 2026.

How a verdict is reached (details in Methods): any failed test gives likely false positive. Warnings, or a test that could not run (for example the density and radius tests when the TIC has no stellar radius), give planet candidate (with caveats), and the reasons name the tests that did not run. Otherwise the verdict is planet candidate (passes all tests). Seven tests use the light curve and one, the centroid test, the target pixels. A clean verdict means the signal is consistent with a planet on the target star: the dip is on the target as far as the pixels can tell, which is to within about 9″ at best, and an eclipsing binary closer to the target than that is not excluded.

TOI-1059.01

• [pass] odd_even: odd depth 24609±252 ppm vs even 24467±215 ppm: 0.4σ difference (uncertainties include the 712 ppm scatter between transits)
• [pass] secondary: no significant eclipse at phase 0.5 (89±86 ppm, 1.0σ); a 499 ppm dip at phase 0.60 (5.9σ) comes from a single orbit and is not counted
• [warn] shape: intermediate: ingress+egress = 0.79 of the duration; posterior P(grazing) = 1.00
• [warn] density: transit-implied ρ* = 2.46 ρ☉ vs catalog 1.05 ρ☉ (ratio 2.35, 3.3σ)
• [fail] radius: companion radius 4.94 R_Jup
• [pass] coverage: 19 of 19 transits with data are fully covered (inside and on both sides)
• [pass] rotation: period is not near the rotation period (10.24 d) or its multiples
• [pass] centroid: the dip is 3.1″ from the target (0.7σ, 4 sectors); stars within 9″ of it cannot be excluded, and no cataloged star there is bright enough to cause it

TOI-4543.01

• [pass] odd_even: odd depth 4356±153 ppm vs even 4460±176 ppm: 0.4σ difference
• [pass] secondary: no significant eclipse at phase 0.5 (-52±80 ppm, -0.6σ)
• [pass] shape: intermediate: ingress+egress = 0.52 of the duration; posterior P(grazing) = 0.00
• [n/a] density: no fitted or catalog density
• [n/a] radius: no stellar radius
• [pass] coverage: 7 of 8 transits with data are fully covered (inside and on both sides)
• [n/a] rotation: no clear rotational modulation
• [pass] centroid: the dip is 1.2″ from the target (0.1σ, 2 sectors); stars within 9″ of it cannot be excluded, and no cataloged star there is bright enough to cause it
• not tested: density, radius, so the verdict rests on the other tests

TOI-4597.01

• [pass] odd_even: odd depth 7207±350 ppm vs even 7401±391 ppm: 0.4σ difference
• [pass] secondary: no significant eclipse at phase 0.5 (2±192 ppm, 0.0σ)
• [pass] shape: U-shaped: ingress+egress = 0.19 of the duration; posterior P(grazing) = 0.00
• [warn] density: transit-implied ρ* = 1.52 ρ☉ vs catalog 0.47 ρ☉ (ratio 3.24, 3.8σ)
• [pass] radius: companion radius 1.17 R_Jup
• [pass] coverage: 9 of 9 transits with data are fully covered (inside and on both sides)
• [n/a] rotation: no clear rotational modulation
• [pass] centroid: the dip is 2.8″ from the target (0.6σ, 2 sectors); stars within 9″ of it cannot be excluded, and no cataloged star there is bright enough to cause it

TOI-1019.01

• [pass] odd_even: odd depth 20783±57 ppm vs even 20846±59 ppm: 0.8σ difference
• [pass] secondary: no significant eclipse at phase 0.5 (21±31 ppm, 0.7σ)
• [pass] shape: U-shaped: ingress+egress = 0.45 of the duration; posterior P(grazing) = 0.00
• [pass] density: transit-implied ρ* = 0.45 ρ☉ vs catalog 0.47 ρ☉ (ratio 0.97, 0.2σ)
• [pass] radius: companion radius 2.18 R_Jup
• [pass] coverage: 39 of 40 transits with data are fully covered (inside and on both sides)
• [n/a] rotation: no clear rotational modulation
• [pass] centroid: the dip is 0.1″ from the target (0.0σ, 4 sectors); stars within 9″ of it cannot be excluded, and no cataloged star there is bright enough to cause it

TOI-1717.01

• [pass] odd_even: odd depth 8541±542 ppm vs even 8723±626 ppm: 0.2σ difference (uncertainties include the 1877 ppm scatter between transits)
• [pass] secondary: no significant eclipse at phase 0.5 (-115±215 ppm, -0.5σ)
• [pass] shape: U-shaped: ingress+egress = 0.46 of the duration; posterior P(grazing) = 0.00
• [pass] density: transit-implied ρ* = 0.52 ρ☉ vs catalog 0.55 ρ☉ (ratio 0.95, 0.2σ)
• [pass] radius: companion radius 1.25 R_Jup
• [pass] coverage: 21 of 21 transits with data are fully covered (inside and on both sides)
• [pass] rotation: period is not near the rotation period (0.30 d) or its multiples
• [pass] centroid: the dip is 0.9″ from the target (0.1σ, 4 sectors); stars within 10″ of it cannot be excluded: TIC 743431875 (Tmag 13.4, 3″) could cause it

Each candidate’s report folder (results/candidates/TOI-.../) contains the search, fit, corner, four-panel vetting and centroid figures that the verdict is based on.

What the verdicts rest on

The selection favors deep, frequent transits on bright stars, and all five candidates are 0.5–2.5 % deep in the TOI catalog: the range of giant planets, and of the eclipsing binaries that imitate them. The pipeline recovered every one at the catalog period, and the centroid test puts all five dips on their target stars, within 3.1″ (0.7σ or less).

• TOI-1717.01 passes every test. A companion of 1.25 R_J on a 4.05-day orbit around a 6,578 K star, with equal odd and even depths, no secondary eclipse, a U-shaped transit and a transit-implied density that matches the catalog (0.52 against 0.55 ρ☉). The star varies strongly at 0.30 days (1,136 ppm), far from the orbital period but faster than the 0.75-day detrending window can follow, and single transits scatter in depth by 1,877 ppm. The odd/even test allows for that scatter (0.2σ). The centroid test puts the dip 0.9″ from the star (0.1σ, four sectors), but it cannot exclude TIC 743431875, a star of magnitude 13.4 only 3″ away that is bright enough to cause the dip.
• TOI-1019.01 passes every test. Measured against the flux around each transit, its odd and even transits are equally deep (20,783 ± 57 and 20,846 ± 59 ppm, 0.8σ). The first run measured every transit against one reference level for the whole light curve and found a difference of 1.1 % (3.7σ), enough to label it a likely false positive; that difference came from the flux level around the transits, not from the transits. The density matches the catalog (0.45 against 0.47 ρ☉), and the centroid test puts the dip 0.1″ from the star (0.0σ, four sectors, each with the dip at S/N 100 or more). The companion, 2.18 R_J, would be larger than almost every known planet, and the light curve cannot say whether it is one or a small star.
• TOI-4543.01 passes the tests that could run, with a caveat because two could not. The TIC lists no radius or density for this bright star (TESS magnitude 6.8), so neither the density nor the radius test ran, and the companion’s size is unknown. The transit implies a star of only 0.07 ρ☉, far less dense than a dwarf, as for an evolved star. Only two sectors of 2-minute data exist, and in both the dip is on the star (1.2″, 0.1σ). The first run called this “passes all tests”, which overstated the evidence.
• TOI-4597.01 passes with a caveat. A 1.17 R_J companion on a 4.67-day orbit around a 7,712 K star, with no odd/even difference and no secondary eclipse. But the transit implies a star 3.2 times denser than the catalog value (3.8σ), which earns a density warning. An eccentric orbit can do that (see the vetting page), and so can an error in the catalog’s stellar radius. Two sectors, in which the dip is on the star (2.8″, 0.6σ).
• TOI-1059.01 is labeled a likely false positive by the radius test: the implied companion is 4.94 R_J. The dip is 2.5 % deep, and the fit puts it on a grazing orbit (b = 1.32, probability of grazing 1.00, which also earns a shape warning). For a grazing transit the size is poorly constrained, but even the lower end of the 68 % interval of the radius, about 2.9 R_J (55.1, −23/+27 R⊕), is above the 2.5 R_J limit, and the centroid test puts the dip on TOI-1059 itself (3.1″, 0.7σ). The search also found two more signals in this star’s light curve, which the vetting rejects: at 78.1 days, two dips, only one of them covered by data on both sides (odd/even, density and radius tests), and at 52.1 days, three dips that all sit at the edges of data segments (odd/even and coverage tests) and 20″ from the target, where no cataloged star is bright enough to cause them (centroid test).

One highly ranked TOI, TOI-651.01, was left out. Its SPOC light curves are filed under TIC 72090499, a separate catalog entry at the same position, not under the TOI’s TIC 72090501, and the pipeline never uses another target’s light curves.

These verdicts come from the TESS data alone. The centroid test puts all five dips on their stars, but it cannot rule out a binary within about 9″ of the target, such as the star 3″ from TOI-1717, and none of the verdicts replaces the TESS Follow-up Observing Program’s photometry, imaging and spectroscopy.