We investigate the mechanics ofa new double chain-model of DNA. The model consists of two long elastic homogeneous strands (or rods), which represent two polynucleotide chains of the DNA molecule, connected with each ...We investigate the mechanics ofa new double chain-model of DNA. The model consists of two long elastic homogeneous strands (or rods), which represent two polynucleotide chains of the DNA molecule, connected with each other by an elastic membrane (or some linear springs) representing the hydrogen bonds between the base pairs of the two chains. The nonlinear dynamical equations are derived and some solitary wave solutions are discussed. By the way,our model can be regarded as the mechanical model of cubic nonlinear Klein-Gordon equation and 4-field equation.展开更多
大语言模型(Large Language Model, LLM),尤其是一些预训练的深度神经网络,拥有强大的表征学习能力,不仅能学习典型的软件缺陷,还能够深度分析软件隐含的缺陷,给出风险提示和改进建议。大模型的应用,使得测试方法和流程更加智能化,且能...大语言模型(Large Language Model, LLM),尤其是一些预训练的深度神经网络,拥有强大的表征学习能力,不仅能学习典型的软件缺陷,还能够深度分析软件隐含的缺陷,给出风险提示和改进建议。大模型的应用,使得测试方法和流程更加智能化,且能够发现之前难以捕捉的复杂缺陷。航天器软件以其独特性、专业性、复杂性、高可靠性、高安全性的特点,需要通过专业的方法训练专业的大模型来满足航天嵌入式软件测试的需要。研究表明,大模型结合有效的测试方法形成的完备思维链(Chain of Thought, Cot)可以有效提高软件缺陷检出率。将测试专家的思维和大模型相结合,用专家思维链引导大模型查找软件缺陷,在提示词中增加测试专家的思维链可以有效提升软件缺陷检出效率,保证软件质量,显著提升测试团队的整体专业水平。展开更多
文摘We investigate the mechanics ofa new double chain-model of DNA. The model consists of two long elastic homogeneous strands (or rods), which represent two polynucleotide chains of the DNA molecule, connected with each other by an elastic membrane (or some linear springs) representing the hydrogen bonds between the base pairs of the two chains. The nonlinear dynamical equations are derived and some solitary wave solutions are discussed. By the way,our model can be regarded as the mechanical model of cubic nonlinear Klein-Gordon equation and 4-field equation.
文摘大语言模型(Large Language Model, LLM),尤其是一些预训练的深度神经网络,拥有强大的表征学习能力,不仅能学习典型的软件缺陷,还能够深度分析软件隐含的缺陷,给出风险提示和改进建议。大模型的应用,使得测试方法和流程更加智能化,且能够发现之前难以捕捉的复杂缺陷。航天器软件以其独特性、专业性、复杂性、高可靠性、高安全性的特点,需要通过专业的方法训练专业的大模型来满足航天嵌入式软件测试的需要。研究表明,大模型结合有效的测试方法形成的完备思维链(Chain of Thought, Cot)可以有效提高软件缺陷检出率。将测试专家的思维和大模型相结合,用专家思维链引导大模型查找软件缺陷,在提示词中增加测试专家的思维链可以有效提升软件缺陷检出效率,保证软件质量,显著提升测试团队的整体专业水平。
