TokenAdapt initializes new token embeddings by blending subword-decomposition and semantic-neighbor estimates, and achieves lower zero-shot perplexity after tokenizer replacement than ReTok and TransTokenizer, while supertoken learning improves text compression.
Entanglement generation in capacitively coupled Transmon-cavity system
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
abstract
In this paper, the higher energy levels of the transmon qubit are taken into consideration to investigate the continuous variable entanglement generation between the transmon qubit and the single-mode cavity. Based on the framework of cavity quantum electrodynamics, we show the entanglement generation depends on the the driving field intensity, coupling strength, cavity field frequency, and qubit frequency. The numerical results show that strong entanglement can be generated by properly tuning these parameters. It is our hope that the results presented in this paper may lead to a better understanding of quantum entanglement generation in cavity QED system and provide new perspectives for further research in quantum information processing.
fields
cs.CL 1years
2025 1verdicts
REJECT 1representative citing papers
citing papers explorer
-
Achieving Tokenizer Flexibility in Language Models through Heuristic Adaptation and Supertoken Learning
TokenAdapt initializes new token embeddings by blending subword-decomposition and semantic-neighbor estimates, and achieves lower zero-shot perplexity after tokenizer replacement than ReTok and TransTokenizer, while supertoken learning improves text compression.