Channel Stability Analysis by One-Way Fluid Structure Interaction: A Case Study in China

Abstract

The wide adoption of Large language models (LLMs) makes their dependability a pressing concern. Detection of errors is the first step to mitigating their impact on a system and thus, efficient error detection for LLMs is an important issue. In many settings, the LLM is considered as a black box with no access to the internal nodes; this prevents the use of many error detection schemes that need access to the model's internal nodes. An interesting observation is that the output of LLMs in error-free operation should be valid and normal text. Therefore, when the text is not valid or differs significantly from normal text, it is likely that there is an error. Based on this observation we propose to perform Concurrent Linguistic Error Detection (CLED); this scheme extracts some linguistic features of the text generated by the LLM and feeds them to a concurrent classifier that detects errors. Since the proposed error detection mechanism only relies on the outputs of the model, then it can be used on LLMs in which there is no access to the internal nodes. The proposed CLED scheme has been evaluated on the T5 model when used for news summarization and on the OPUS-MT model when used for translation. In both cases, the same set of linguistic features has been used for error detection to illustrate the applicability of the proposed scheme beyond a specific case. The results show that CLED can detect most of the errors at a low overhead penalty. The use of the concurrent classifier also enables a trade-off between error detection effectiveness and its associated overhead, so providing flexibility to a designer.

Stability on the Excavation Surface of Submarine Shield Tunnel Considering the Fluid–Solid Coupling Effect and the Equivalent Layer

“…However, the ultimate support ratio at the excavation face decreases with the increase in the internal friction angle of the soil. These conclusions are basically consistent with the conclusions of Zhang et al [33], Kang, et al [36], and Zhu et al [37] Most of the articles have not discussed the thickness of the equivalent layer, which is related to the amount of grouting during construction and the extent of the disturbance of the surrounding soil behind the lining wall. Therefore, it is necessary to avoid over-excavation and disturbing the surrounding soil during tunnel construction to reduce the thickness of the equivalent layer in order to control the displacement of the ground surface.…”

“…When the excavation face is destabilized, the sliding body in the front of the excavation is wedge-shaped, and there is chimney-shaped damage on the whole area. Zhu et al [37] investigated the method of determining the ultimate supporting forces of the excavation face in the presence of groundwater and the damage modes of the excavation face in sandy soil stratum through FLAC3D software. The results obtained in their paper are approximately the same as the present one.…”

Data-Driven Recovery Potential Analysis and Modeling for Batteries Recovery Operations in Electric Bicycle Industry

“…In China over 85% of the raw materials of secondary lead are from waste lead-acid batteries, and 50% of secondary lead is used to produce storage batteries (Zhu S.R. 2002) [28]. Therefore, to be able to reasonably, scientifically, and accurately predict the generation amount of waste lead-acid batteries and estimate the amount of secondary lead and other renewable resources recycled from waste lead-acid batteries is of great significance to the waste metal resource recovery, ecological environment protection, metal resource safety, and sustainable development of the (secondary) lead industry of China.…”

Analysis of Binding Modes between Three Perfluorosulfonates and GPER Based on Computational Simulation and Multiple Spectral Methods

“…Perfluoroalkyl sulfonic acid (PFSAs) is widely recognized as one of the most significant chemical products of the 20th century. Fluoride usage is prevalent in both the production and consumption of the organic industry [1,2]. PFSAs find extensive application in various domains, including surface antifouling agents, foam fire extinguishing agents, paint additives, pesticides, and herbicides, because of their desirable oil and water-repellent properties.…”