In recent years, A variety of methods based on Transformer and state space model (SSM) architectures have been proposed, advancing foundational DNA language models. However, there is a lack of comparison between these recent approaches and the classical architecture—convolutional networks (CNNs)—on foundation model benchmarks. This raises the question: are CNNs truly being surpassed by these recent approaches based on transformer and SSM architectures? In this paper, we develop a simple but well-designed CNN-based method, termed ConvNova. ConvNova identifies and proposes three effective designs: 1) dilated convolutions, 2) gated convolutions, and 3) a dual-branch framework for gating mechanisms. Through extensive empirical experiments, we demonstrate that ConvNova significantly outperforms recent methods on more than half of the tasks across several foundation model benchmarks. For example, in histone-related tasks, ConvNova exceeds the second-best method by an average of 5.8\%, while generally utilizing fewer parameters and enabling faster computation. In addition, the experiments observed findings that may be related to biological characteristics. This indicates that CNNs are still a strong competitor compared to Transformers and SSMs. We anticipate that this work will spark renewed interest in CNN-based methods for DNA foundation models.

A simultaneous packet transmission (SPT) technique is proposed to increase the throughput of a communication system. A SPT scheme employs multiple levels and employs multiple streams at each level above the first level. Every stream of the SPT has its assigned partition on the overall constellation. Symbols of an N ‐level SPT are formed by taking , bits from a selected stream of every level. The SPT scheme can employ higher code rates on streams at higher levels to increase the system throughput. Numerical results show that the SPT offers 15–30% throughput enhancement in turbo‐coded LTE systems.

Traditionally, concatenation of codes is used to increase the coding gain at the expense of the code rate and the throughput. In contrast, throughput enhancing concatenated codes (TECCs) are constructed here aiming to increase the throughput. In a TECC, the coded sequence of a code C transmitted through the channel is altered according to the coded bits of a second coded sequence of an implicit code . In this study, TECCs select one bit in every segment of n coded bits of C based on coded bits of , and flip that selected coded bit of C before transmission. Even though the coded bits of are not transmitted explicitly over the channel, the information of coded bits of is transferred to the destination through the altered bits which can be recovered by exploiting the extrinsic information of explicitly coded bits. Using iterative decoding between codes C and , the receiver can decode the coded bits of both the explicit code C and the implicit code jointly. TECCs that can increase the throughput of C by 25–37.5% without sacrificing the performance of C are reported, however, at the expense of increased complexity.

Even in the field of biology, the principle of how knowledge can be efficiently utilized in a neural network has not been solved perfectly. This paper discusses a new model based on the structure of the V1 area in biological visual system for knowledge matching. The contour of the object is stored as knowledge in the form of chain code. During the matching process, the chain code is presented to control the dynamics of the neuron in a V1-similar structure neural network. Cooperating with the dynamic weights, active neurons will reconstruct the object's contour at the place where the object is in the visual field. This model is an exploration of how knowledge is represented and utilized in the brain.