Search for: All records

Transformer models have been widely investigated in different domains by providing long-range dependency handling and global contextual awareness, driving the development of popular AI applications such as ChatGPT, Gemini, and Alexa. State Space Models (SSMs) have emerged as strong contenders in the field of sequential modeling, challenging the dominance of Transformers. SSMs incorporate a selective mechanism that allows for dynamic parameter adjustment based on input data, enhancing their performance. However, this mechanism also comes with increasing computational complexity and bandwidth demands, posing challenges for deployment on resource-constraint mobile devices. To address these challenges without sacrificing the accuracy of the selective mechanism, we propose a sparse learning framework that integrates architecture-aware compiler optimizations. We introduce an end-to-end solution–C 4 n kernel sparsity, which prunes n elements from every four contiguous weights, and develop a compiler-based acceleration solution to ensure execution efficiency for this sparsity on mobile devices. Based on the kernel sparsity, our framework generates optimized sparse models targeting specific sparsity or latency requirements for various model sizes. We further leverage pruned weights to compensate for the remaining weights, enhancing downstream task performance. For practical hardware acceleration, we propose C 4 n -specific optimizations combined with a layout transformation elimination strategy. This approach mitigates inefficiencies arising from fine-grained pruning in linear layers and improves performance across other operations. Experimental results demonstrate that our method achieves superior task performance compared to other semi-structured pruning methods and achieves up-to 7→ speedup compared to llama.cpp framework on mobile devices. 

Abstract The stress field perturbation caused by magmatic intrusions within volcanic systems induces strain in the surrounding region. This effect results in the opening and closing of microcracks in the vicinity of the intrusion, which can affect regional seismic velocities. In late November 2023, we deployed a distributed acoustic sensing interrogator to convert an existing 100‐km telecommunication fiber‐optic cable along the coast of Iceland's Reykjanes peninsula into a dense seismic array, which has run continuously. Measuring changes in surface wave moveout with ambient noise cross‐correlation, we observe up to 2% changes in Rayleigh wave phase velocity following eruptions in the peninsula's 2023–2024 sequence that are likely associated with magmatic intrusions into the eruption‐feeding dike. We apply a Bayesian inversion to compute the posterior distribution of potential dike opening models for each eruption by considering measurements for varying channel pairs and frequency bands, and assuming this velocity change is tied to volumetric strain associated with dike‐opening. Our results are in agreement with those based on geodetic measurement and provide independent constraints on the depth of the dike, demonstrating the viability of this novel inversion and new volcano monitoring directions through fiber sensing. 

Abstract Although typically used to measure dynamic strain from seismic and acoustic waves, Rayleigh‐based distributed acoustic sensing (DAS) is also sensitive to temperature, offering longer range and higher sensitivity to small temperature perturbations than conventional Raman‐based distributed temperature sensing. Here, we demonstrate that ocean‐bottom DAS can be employed to study internal wave and tide dynamics in the bottom boundary layer, a region of enhanced ocean mixing but scarce observations. First, we show temperature transients up to about 4 K from a power cable in the Strait of Gibraltar south of Spain, associated with passing trains of internal solitary waves in water depth <200 m. Second, we show the propagation of thermal fronts associated with the nonlinear internal tide on the near‐critical slope of the island of Gran Canaria, off the coast of West Africa, with perturbations up to about 2 K at 1‐km depth and 0.2 K at 2.5‐km depth. With spatial averaging, we also recover a signal proportional to the barotropic tidal pressure, including the lunar fortnightly variation. In addition to applications in observational physical oceanography, our results suggest that contemporary chirped‐pulse DAS possesses sufficient long‐period sensitivity for seafloor geodesy and tsunami monitoring if ocean temperature variations can be separated. 

We report the confirmation and mass determination of three hot Jupiters discovered by the Transiting Exoplanet Survey Satellite (TESS) mission: HIP 65Ab (TOI-129, TIC-201248411) is an ultra-short-period Jupiter orbiting a bright ( V = 11.1 mag) K4-dwarf every 0.98 days. It is a massive 3.213 ± 0.078  M J planet in a grazing transit configuration with an impact parameter of b = 1.17 −0.08 +0.10 . As a result the radius is poorly constrained, 2.03 −0.49 +0.61 R J . The planet’s distance to its host star is less than twice the separation at which it would be destroyed by Roche lobe overflow. It is expected to spiral into HIP 65A on a timescale ranging from 80 Myr to a few gigayears, assuming a reduced tidal dissipation quality factor of Q s ′ = 10 7 − 10 9 . We performed a full phase-curve analysis of the TESS data and detected both illumination- and ellipsoidal variations as well as Doppler boosting. HIP 65A is part of a binary stellar system, with HIP 65B separated by 269 AU (3.95 arcsec on sky). TOI-157b (TIC 140691463) is a typical hot Jupiter with a mass of 1.18 ± 0.13  M J and a radius of 1.29 ± 0.02  R J . It has a period of 2.08 days, which corresponds to a separation of just 0.03 AU. This makes TOI-157 an interesting system, as the host star is an evolved G9 sub-giant star ( V = 12.7). TOI-169b (TIC 183120439) is a bloated Jupiter orbiting a V = 12.4 G-type star. It has a mass of 0.79 ±0.06  M J and a radius of 1.09 −0.05 +0.08 R J . Despite having the longest orbital period ( P = 2.26 days) of the three planets, TOI-169b receives the most irradiation and is situated on the edge of the Neptune desert. All three host stars are metal rich with [Fe / H] ranging from 0.18 to0.24.