The pulsar glitches are generally assumed to be an apparent manifestation of the superfluid interior of the neutron stars. Most of them were discovered and extensively studied by continuous monitoring in the radio wavelengths. The Fermi-LAT space telescope has made a revolution uncovering a large population of gamma-ray pulsars. In this paper we suggest to employ these observations for the searches of new glitches. We develop the method capable of detecting step-like frequency change associated with glitches in a sparse gamma-ray data. It is based on the calculations of the weighted H-test statistics and glitch identification by a convolutional neural network. The method demonstrates high accuracy on the Monte Carlo set and will be applied for searches of the pulsar glitches in the real gamma-ray data in the future works.

Observations by Fermi LAT enabled us to explore the population of non-recycled gamma-ray pulsars with the set of 89 objects. It was recently noted that there are apparent differences in properties of radio-quiet and radio-loud subsets. In particular, average observed radio-loud pulsar is younger than radio-quiet one and is located at smaller galactic latitude. Even so, the analysis based on the full list of pulsars may suffer from selection effects. Namely, most of radio-loud pulsars are first discovered in the radio-band, while radio-quiet ones are found using the gamma-ray data. In this work we perform a blind search for gamma-ray pulsars using the Fermi LAT data alone using all point sources from 3FGL catalog as the candidates. Unlike preceding blind search, the present catalog is constructed with novel semi-coherent method and covers the full range of characteristic ages down to 1 kyr. The search resulted in the catalog of 40 non-recycled pulsars, 26 of which are radio-quiet. There are no statistically significant differences in age and galactic latitude distributions for the radio-loud and radio-quiet pulsars, while the rotation period distributions are marginally different with $2.4\sigma$ pre-trial statistical significance. The fraction of radio-quiet pulsars is estimated as $\epsilon_{RQ}=63\pm 8\%$. The results are in agreement with the predictions of the outer magnitosphere models, while the Polar cap models are disfavored.

The Fermi Large Area Telescope (LAT) has observed more than a hundred of gamma-ray pulsars, about one third of which are radio-quiet, i.e. not detected at radio frequencies. The most of radio-loud pulsars are detected by Fermi LAT by using the radio timing models, while the radio-quiet ones are discovered in a blind search. The difference in the techniques introduces an observational selection bias and, consequently, the direct comparison of populations is complicated. In order to produce an unbiased sample, we perform a blind search of gamma-ray pulsations using Fermi-LAT data alone. No radio data or observations at optical or X-ray frequencies are involved in the search process. We produce a gamma-ray selected catalog of 25 non-recycled gamma-ray pulsars found in a blind search, including 16 radio-quiet and 9 radio-loud pulsars. This results in the direct measurement of the fraction of radio-quiet pulsars $\varepsilon_{RQ} = 64\pm 10\%$, which is in agreement with the existing estimates from the population modeling in the outer magnetosphere model. The Polar cap models are disfavored due to a lower expected fraction and the prediction of age dependence. The age, gamma-ray energy flux, spin-down luminosity and sky location distributions of the radio-loud and radio-quiet pulsars from the catalog do not demonstrate any statistically significant difference. The results indicate that the radio-quiet and radio-loud pulsars belong to one and the same population. The catalog shows no evidence for the radio beam evolution.