Efficient data selection is crucial to accelerate the pretraining of language model (LMs). While various methods have been proposed to enhance data efficiency, limited research has addressed the inherent conflicts between these approaches to achieve optimal data selection for LM pretraining. To tackle this problem, we propose a multi-actor collaborative data selection mechanism: each data selection method independently prioritizes data based on its criterion and updates its prioritization rules using the current state of the model, functioning as an independent actor for data selection; and a console is designed to adjust the impacts of different actors at various stages and dynamically integrate information from all actors throughout the LM pretraining process. We conduct extensive empirical studies to evaluate our multi-actor framework. The experimental results demonstrate that our approach significantly improves data efficiency, accelerates convergence in LM pretraining, and achieves an average relative performance gain up to 10.5% across multiple language model benchmarks compared to the state-of-the-art methods. Code and checkpoints are publicly released at https://github.com/Relaxed-System-Lab/multi-actor-data-selection.

Xinlin Zhuang, Jiahui Peng, Ren Ma, Yinfan Wang, Tianyi Bai, Xingjian Wei, Qiu Jiantao, Chi Zhang, Ying Qian, Conghui He. Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2025.

Continuous-flow microfluidic chips are multilayered miniaturized platforms to manipulate small volumes of fluids with valves. There are two types of channels on a chip: flow channels for the reaction of fluids, and control channels for the actuation of valves. Multiplexers (MUXes) are essential microfluidic components for individually addressing many flow channels with few control channels. As the integration scale of microfluidic chips increases, the reliability of MUXes becomes a critical concern, as a single defective control channel in a MUX will affect a large part of the flow channels addressed by the MUX. This paper formally analyzes and identifies the design rules for a MUX to tolerate n defective control channels, and model the fault-tolerant MUX (FT-MUX) design problem as a binary constant weight code problem to minimize resource overheads. We demonstrate that FT-MUX improves resource efficiency by up to hundreds of times compared to the conventional fault-tolerant design method. Besides, given no less than 10 control channels, FT-MUX tolerates at least one defective control channel and addresses even more flow channels with equal or fewer resources than a standard MUX. The advantages become more significant as the integration scale increases.

The integration of visual encoders and large language models (LLMs) has driven recent progress in multimodal large language models (MLLMs). However, the scarcity of high-quality instruction-tuning data for vision-language tasks remains a challenge. The current leading paradigm, such as LLaVA, relies on language-only GPT-4 to generate data, which requires pre-annotated image captions and detection bounding boxes, suffering from understanding image details. A practical solution to this problem would be to utilize the available multimodal large language models to generate instruction data for vision-language tasks. However, it's worth noting that the currently accessible MLLMs are not as powerful as their LLM counterparts, as they tend to produce inadequate responses and generate false information. As a solution for addressing the current issue, this paper proposes the Visual Instruction Generation and Correction (VIGC) framework that enables multimodal large language models to generate instruction-tuning data and progressively enhance its quality on-the-fly. Specifically, Visual Instruction Generation (VIG) guides the vision-language model to generate diverse instruction-tuning data. To ensure generation quality, Visual Instruction Correction (VIC) adopts an iterative update mechanism to correct any inaccuracies in data produced by VIG, effectively reducing the risk of hallucination. Leveraging the diverse, high-quality data generated by VIGC, we finetune mainstream models and validate data quality based on various evaluations. Experimental results demonstrate that VIGC not only compensates for the shortcomings of language-only data generation methods, but also effectively enhances the benchmark performance. The models, datasets, and code are available at https://opendatalab.github.io/VIGC

Multi-temporal SAR and Polarimetric SAR (PolSAR) images can provide the scattering change characteristics caused by vegetation growth to help the classifier capture phenological characteristics. To effectively utilize multi-temporal PolSAR data, a triple temporal vision transformer (TriTempoViT) model is proposed to capture correlation from multi-dimensional features. The method uses a three-branch network architecture to extract spatial-temporal, spatial-polarimetric, and temporal-polarization features respectively, and then the features from the three branches will be integrated into the vision transformer (ViT) for information interaction. Additionally, a 3D channel-spatial attention module (3DCSAM) is tailored to automatically weight the importance of the multi-dimensional feature maps. Moreover, a temporal interaction feature extraction module (TIFEM) is designed to comprehensively consider correlations between different temporal sequences. Compared with the recently developed state-of-the-art approach, the proposed method can improve OA of the classification accuracy in a Radarsat-2 dataset from Flevoland by about 0.84%, which proves the effectiveness of the proposed method.