Results 101 to 110 of about 152,555 (310)
The Token Tax: Systematic Bias in Multilingual Tokenization
Tokenization inefficiency imposes structural disadvantages on morphologically complex, low-resource languages, inflating compute resources and depressing accuracy. We evaluate 10 large language models (LLMs) on AfriMMLU (9,000 MCQA items; 5 subjects; 16 African languages) and show that fertility (tokens/word) reliably predicts accuracy.
Jessica M. Lundin +6 more
openaire +2 more sources
Multimodal Data‐Driven Microstructure Characterization
A self‐consistent autonomous workflow for EBSP‐based microstructure segmentation by integrating PCA, GMM clustering, and cNMF with information‐theoretic parameter selection, requiring no user input. An optimal ROI size related to characteristic grain size is identified.
Qi Zhang +4 more
wiley +1 more source
Challenges and Prospects of Crypto-Currency Market Development
The article deals with the research on global crypto-currency market and analysis of its development trends. Special attention was paid to national regulation of crypto-currency and activity connected with it, which is topical today due to introduction ...
O. P. Kultygin +3 more
doaj +1 more source
Microstructure Evolution of a VMnFeCoNi High‐Entropy Alloy After Synthesis, Swaging, and Annealing
The synthesis and processing (rotary swaging and annealing) of the novel VMnFeCoNi alloy is investigated, alongside the estimation of the grain size effect on hardness. Analysis of a wide grain size range of recrystallized microstructures (12–210 µm) reveals a low annealing twin density.
Aditya Srinivasan Tirunilai +6 more
wiley +1 more source
Token Bottleneck: One Token to Remember Dynamics
Deriving compact and temporally aware visual representations from dynamic scenes is essential for successful execution of sequential scene understanding tasks such as visual tracking and robotic manipulation. In this paper, we introduce Token Bottleneck (ToBo), a simple yet intuitive self-supervised learning pipeline that squeezes a scene into a ...
Taekyung Kim 0002 +4 more
openaire +2 more sources
Hydrogen‐Assisted Fracture of Iron‐Based Fe–Ni–Al Alloys
Principal relations and fracture mechanisms of single‐phase and precipitate‐strengthened Fe–Ni–Al alloys subjected to prior electrochemical hydrogen charging are identified. The mechanisms of hydrogen effect on strength and microhardness are discussed, including hydrogen‐induced increase in microhardness and the role of hydrogen in fracture behavior ...
Nataliya Yadzhak +3 more
wiley +1 more source
The temperature dependence of fatigue behavior in nickel‐based superalloys is investigated through high‐resolution measurements of plastic localization. While increasing temperature reduces localization and enhances fatigue performance in René 88DT, Inconel 718 exhibits a sharp degradation at intermediate temperature due to intensified slip ...
M. Calvat +5 more
wiley +1 more source
Fatigue Crack Initiation and Growth in Nanocrystalline Ni at Multiple Length‐Scales
Overview of miniaturized in situ SEM fatigue setup and resultant fatigue crack growth data for nanocrystalline Ni. The presented study focuses on the analysis of fatigue crack growth rate (FCGR) in focused ion beam‐notched microcantilevers prepared from nanocrystalline (NC) Ni as a model material.
Igor Moravcik +7 more
wiley +1 more source
Single-token vs Two-token Blockchain Tokenomics
We study long-term equilibria that arise in the token monetary policy, or tokenomics, design of proof-of-stake (PoS) blockchain systems that engage utility maximizing users and validators. Validators are system maintainers who get rewarded with tokens for performing the work necessary for the system to function properly, while users compete and pay ...
Kiayias, Aggelos; id_orcid 0000-0003-2451-1430 +2 more
openaire +4 more sources

