Large Language Models (LLMs) demonstrate impressive capabilities, yet their outputs often suffer from misalignment with human preferences due to the inadequacy of weak supervision and a lack of fine-grained control. Training-time alignment methods like Reinforcement Learning from Human Feedback (RLHF) face prohibitive costs in expert supervision and inherent scalability limitations, offering limited dynamic control during inference. Consequently, there is an urgent need for scalable and adaptable alignment mechanisms. To address this, we propose W2S-AlignTree, a pioneering plug-and-play inference-time alignment framework that synergistically combines Monte Carlo Tree Search (MCTS) with the Weak-to-Strong Generalization paradigm for the first time. W2S-AlignTree formulates LLM alignment as an optimal heuristic search problem within a generative search tree. By leveraging weak model's real-time, step-level signals as alignment proxies and introducing an Entropy-Aware exploration mechanism, W2S-AlignTree enables fine-grained guidance during strong model's generation without modifying its parameters. The approach dynamically balances exploration and exploitation in high-dimensional generation search trees. Experiments across controlled sentiment generation, summarization, and instruction-following show that W2S-AlignTree consistently outperforms strong baselines. Notably, W2S-AlignTree raises the performance of Llama3-8B from 1.89 to 2.19, a relative improvement of 15.9% on the summarization task.

Training Large Language Models (LLMs) is prohibitively expensive, creating a critical scaling gap where insights from small-scale experiments often fail to transfer to resource-intensive production systems, thereby hindering efficient innovation. To bridge this, we introduce Farseer, a novel and refined scaling law offering enhanced predictive accuracy across scales. By systematically constructing a model loss surface $L(N,D)$, Farseer achieves a significantly better fit to empirical data than prior laws (e.g., \Chinchilla's law). Our methodology yields accurate, robust, and highly generalizable predictions, demonstrating excellent extrapolation capabilities, outperforming Chinchilla's law, whose extrapolation error is 433\% higher. This allows for the reliable evaluation of competing training strategies across all $(N,D)$ settings, enabling conclusions from small-scale ablation studies to be confidently extrapolated to predict large-scale performance. Furthermore, Farseer provides new insights into optimal compute allocation, better reflecting the nuanced demands of modern LLM training. To validate our approach, we trained an extensive suite of approximately 1,000 LLMs across diverse scales and configurations, consuming roughly 3 million NVIDIA H100 GPU hours. To foster further research, we are comprehensively open-sourcing all code, data, results (https://github.com/Farseer-Scaling-Law/Farseer), all training logs (https://wandb.ai/billzid/Farseer?nw=nwuserbillzid), all models used in scaling law fitting (https://huggingface.co/Farseer-Scaling-Law).

Traditional route forwarding generates obvious performance problems and it cannot fulfill the increasing diversity in the number of user accesses and Quality of Service (QoS). It is necessary to investigating an intelligent and reliable routing for online audio/video mixed traffic with high-real time to meet different QoS requirements. In this article, we design a routing scheme based on deep reinforcement learning (DRL) and graph neural networks (GNNs), which could be easily implemented as an application on a controller, named RtDG. Specifically, a network topology is first extracted using GNN to generate high-dimensional feature representations. To obtain QoS utility values comprehensively, we set four parameters, namely, bandwidth, delay, packet loss rate, and delay jitter, and construct weighted formulas using the parameters determined by Bayesian optimization. We use proximal policy optimization (PPO) to make output decisions while adding a KL scatter penalty to the loss function. And then, the controller assigns it to switches via traffic table based on the calculated QoS values. Furthermore, we deploy an overlay network using Mininet and ONOS, enabling optimal pathfinding without changing the existing network architecture. Meanwhile, tests are conducted under background traffic of online audio/video. Extensive simulation results demonstrate that the RtDG can significantly reduce average delay and packet loss rate by 52.21% and 57.83%, compared to traditional routing strategies. Especially under high traffic conditions, it is able to consider the uncertainty during path selection and achieve excellent routing performance.

A major challenge in analysis of huge amounts of ocean data is the complexity of the data and the inherent complexity of ocean dynamic processes. Interactive visual analysis serves as an efficient complementary approach for the detection of various phenomena or patterns, and correlation exploring or comparing multiple variables in researchers daily work. Firstly, this paper presents a basic concept of the ocean data produced from numerous measurement devices or computer simulations. The characteristics of ocean data and the related data processing techniques are also described. Secondly, the main tasks of ocean data analysis are introduced. Based on the main analysis tasks in the field of oceanography, the survey emphasizes related interactive visualization techniques and tools from four aspects: visualization of multiple ocean environmental elements and multivariate analysis , ocean phenomena identification and tracking, patterns or correlation discovery, ensembles and uncertainties exploration. Finally, the opportunities are discussed for future studies.