Search for: All records

This study utilizes the magnetic susceptibility (MS) of sedimentary strata to correlate the Late Devonian Antrim Formation black shale and calcareous mudstone within the Michigan Basin as well as the Antrim with previously published MS profiles from contemporaneous, shale-dominated strata from the Illinois Basin. MS can be used as a proxy for changes in material composition, which is linked to paleoclimate-controlled sediment fluxes and depositional environments. In the Michigan Basin, MS profiles through the basin-margin State Chester Welch 18 and the more basinal Krocker 1-17 cores show that MS patterns correspond to lithostratigraphic units. For some of these units the MS patterns are similar among the cores, though not for all units. Preliminary interpretation is that MS patterns are a result of proximity to sediment source (Acadian Orogeny versus Transcontinental Arch) as well as intrabasinal early diagenetic processes (pyrite). Furthermore, the lithostratigraphic units in these cores may not be chronostratigraphically equivalent. This study also compares the Michigan Basin MS basinal profile (Krocker 1-17 core) with previously published data from the “Bullitt County Core” from Kentucky, in the southern Illinois Basin. Within a biostratigraphic framework, the Michigan and Illinois Basin cores appear to show similar MS patterns. This is possibly because sediment input to these two locations is primarily sourced from the Acadian Orogeny, and the depositional environment and therefore early diagenetic processes, are similar. Future work will combine mineralogical analysis with the MS profiles to decipher the source of magnetic susceptibility, currently hypothesized to be driven by ilmenite concentration. 

AI Forensics is a novel research field that aims at providing techniques, mechanisms, processes, and protocols for an AI failure investigation. In this paper, we pave the way towards further exploring a sub-domain of AI forensics, namely AI model forensics, and introduce AI model ballistics as a subfield inspired by forensic ballistics. AI model forensics studies the forensic investigation process, including where available evidence can be collected, as it applies to AI models and systems. We elaborate on the background and nature of AI model development and deployment, and highlight the fact that these models can be replaced, trojanized, gradually poisoned, or fooled by adversarial input. The relationships and the dependencies of our newly proposed subdomain draws from past literature in software, cloud, and network forensics. Additionally, we share a use-case mini-study to explore the peculiarities of AI model forensics in an appropriate context. Blockchain is discussed as a possible solution for maintaining audit trails. Finally, the challenges of AI model forensics are discussed. 

Abstract P-type point contact (PPC) HPGe detectors are a leading technology for rare event searches due to their excellent energy resolution, low thresholds, and multi-site event rejection capabilities. We have characterized a PPC detector’s response to $$\alpha $$ α particles incident on the sensitive passivated and p $$^+$$ + surfaces, a previously poorly-understood source of background. The detector studied is identical to those in the Majorana Demonstrator experiment, a search for neutrinoless double-beta decay ( $$0\nu \beta \beta $$ 0 ν β β ) in $$^{76}$$ 76 Ge. $$\alpha $$ α decays on most of the passivated surface exhibit significant energy loss due to charge trapping, with waveforms exhibiting a delayed charge recovery (DCR) signature caused by the slow collection of a fraction of the trapped charge. The DCR is found to be complementary to existing methods of $$\alpha $$ α identification, reliably identifying $$\alpha $$ α background events on the passivated surface of the detector. We demonstrate effective rejection of all surface $$\alpha $$ α events (to within statistical uncertainty) with a loss of only 0.2% of bulk events by combining the DCR discriminator with previously-used methods. The DCR discriminator has been used to reduce the background rate in the $$0\nu \beta \beta $$ 0 ν β β region of interest window by an order of magnitude in the Majorana Demonstrator   and will be used in the upcoming LEGEND-200 experiment.