Search for: All records

The emergence of seemingly similar representations across tasks and neural architectures suggests that convergent properties may underlie sophisticated behavior. One form of representation that seems particularly fundamental to reasoning in many artificial (and perhaps natural) networks is the formation of world models, which decompose observed task structures into re-usable perceptual primitives and task-relevant relations. In this work, we show that auto-regressive transformers tasked with solving mazes learn to linearly represent the structure of mazes, and that the formation of these representations coincides with a sharp increase in generalization performance. Furthermore, we find preliminary evidence for Adjacency Heads which may play a role in computing valid paths through mazes. 

Abstract Radiation measurement relies on pulse detection, which can be performed using various configurations of high-speed analog-to-digital converters (ADCs) and field-programmable gate arrays (FPGAs). For optimal power consumption, design simplicity, system flexibility, and the availability of DSP slices, we consider the Radio Frequency System-on-Chip (RFSoC) to be a more suitable option than traditional setups. To this end, we have developed custom RFSoC-based electronics and verified its feasibility. The ADCs on RFSoC exhibit a flat frequency response of 1–125 MHz. The root-mean-square (RMS) noise level is 2.1 ADC without any digital signal processing. The digital signal processing improves the RMS noise level to 0.8 ADC (input equivalent 40 μV_rms). Baseline correction via digital signal processing can effectively prevent photomultiplier overshoot after a large pulse. Crosstalk between all channels is less than -55 dB. The measured data transfer speed can support up to 32 kHz trigger rates (corresponding to 750 Mbps). Overall, our RFSoC-based electronics are highly suitable for pulse detection, and after some modifications, they will be employed in the Kamioka Liquid Scintillator Anti-Neutrino Detector (KamLAND). 

Abstract The electron antineutrino flux limits are presented for the brightest gamma-ray burst (GRB) of all time, GRB221009A, over a range of 1.8–200 MeV using the Kamioka Liquid Scintillator Antineutrino Detector. Using multiple time windows ranging from minutes to days surrounding the event to search for electron antineutrinos coincident with the GRB, we set an upper limit on the flux under the assumption of several power-law neutrino source spectra, with power-law indices ranging from 1.5 to 3 in steps of 0.5. No excess was observed in any time windows ranging from seconds to days around the event trigger timeT₀. For a power-law index of 2 and a time window ofT₀ ±  500 s, a flux upper limit of 2.34  ×  10⁹cm⁻²was calculated. The limits are compared to the results presented by IceCube. 

Particle dark matter could belong to a multiplet that includes an electrically charged state. WIMP dark matter (χ0) accompanied by a negatively charged excited state (χ−) with a small mass difference (e.g. < 20 MeV) can form a bound-state with a nucleus such as xenon. This bound-state formation is rare and the released energy is O(1−10) MeV depending on the nucleus, making large liquid scintillator detectors suitable for detection. We searched for bound-state formation events with xenon in two experimental phases of the KamLAND-Zen experiment, a xenon-doped liquid scintillator detector. No statistically significant events were observed. For a benchmark parameter set of WIMP mass mχ0=1 TeV and mass difference Δm=17 MeV, we set the most stringent upper limits on the recombination cross section times velocity 〈σv〉 and the decay-width of χ− to 9.2×10−30cm3/s and 8.7×10−14 GeV, respectively at 90% confidence level. 

Abstract Observation of anti-neutrinos emitted from radioactive isotopes inside Earth(geo-neutrinos) brings direct information on the Earth’s chemical composition and its heat balance, which strongly relate to the Earth’s dynamics. To date, two experiments (KamLAND and Borexino) have measured geo-neutrinos and constrained the range of acceptable models for the Earth’s chemical composition, but distinguishing the mantle flux by land-based detectors is challenging as the crust signal is about 70% of the total anti-neutrino flux. Given the oceanic crust is thinner and has lower concentration of radioactive elements than continental crust, geo-neutrino detector in the ocean, Ocean Bottom Detector (OBD), makes it sensitive to geo-neutrinos originating from the Earth’s mantle. Our working group was jointly constructed from interdisciplinary communities in Japan which include particle physics, geoscience, and ocean engineering. We have started to work on technological developments of OBD. We are now developing a 20 kg prototype liquid scintillator detector. This detector will undergo operation deployment tests at 1 km depth seafloor in 2022.