Recent advancements have increasingly focused on leveraging large language models (LLMs) to construct autonomous agents for complex problem-solving tasks. However, existing approaches predominantly employ a single-agent framework to generate search branches and estimate rewards during Monte Carlo Tree Search (MCTS) planning. This single-agent paradigm inherently limits exploration capabilities, often resulting in insufficient diversity among generated branches and suboptimal planning performance.

Multi-agent reinforcement learning (MARL) has emerged as a powerful framework for modeling autonomous agents that independently optimize their individual objectives. However, in mixed-motive MARL environments, rational self-interested behaviors often lead to collectively suboptimal outcomes situations commonly referred to as social dilemmas. A key challenge in addressing social dilemmas lies in accurately quantifying and representing them in a numerical form that captures how self-interested agent behaviors impact social welfare. To address this challenge, \textit{externalities} in the economic concept is adopted and extended to denote the unaccounted-for impact of one agent's actions on others, as a means to rigorously quantify social dilemmas.

Large language models (LLMs) excel across various tasks, but standard first-order (FO) fine-tuning demands considerable memory, significantly limiting real-world deployment. Recently, zeroth-order (ZO) optimization stood out as a promising memory-efficient training paradigm, avoiding backward passes and relying solely on forward passes for gradient estimation, making it attractive for resource-constrained scenarios. However, ZO method lags far behind FO method in both convergence speed and accuracy. To bridge the gap, we introduce a novel layer-wise divergence analysis that uncovers the distinct update pattern of FO and ZO optimization. Aiming to resemble the learning capacity of FO method from the findings, we propose \textbf{Di}vergence-driven \textbf{Z}eroth-\textbf{O}rder (\textbf{DiZO}) optimization. DiZO conducts divergence-driven layer adaptation by incorporating projections to ZO updates, generating diverse-magnitude updates precisely scaled to layer-wise individual optimization needs. Our results demonstrate that DiZO significantly reduces the needed iterations for convergence without sacrificing throughput, cutting training GPU hours by up to 48\% on various datasets. Moreover, DiZO consistently outperforms the representative ZO baselines in fine-tuning RoBERTa-large, OPT-series, and Llama-series on downstream tasks and, in some cases, even surpasses memory-intensive FO fine-tuning. Our code is released at \url{https://github.com/Skilteee/DiZO}.

Modern foundation models are trained on diverse datasets to enhance generalization across tasks and domains. A central challenge in this process is determining how to effectively mix and sample data from multiple sources. This naturally leads to a multi-task learning (MTL) perspective. While prior work in MTL has emphasized mitigating gradient conflicts, we observe that large-scale pretraining scenarios--such as multilingual or multi-domain training--often exhibit little to no gradient conflict. Motivated by this observation, we propose $\textbf{PiKE}$ ($\textbf{P}$ositive gradient $\textbf{i}$nteraction-based $\textbf{K}$-task weights $\textbf{E}$stimator), an adaptive data mixing algorithm that dynamically adjusts sampling weights during training. PiKE exploits non-conflicting gradient interactions to minimize a near-tight upper bound on the average loss decrease at each step, while incurring negligible computational overhead. We provide theoretical convergence guarantees and show that PiKE outperforms static and non-adaptive mixing baselines. Furthermore, we extend PiKE to promote balanced learning across tasks. Extensive experiments on large-scale language model pretraining confirm that PiKE achieves faster convergence and improved downstream performance compared to existing approaches.

We propose Modality-Aware Sharpness-Aware Minimization (M-SAM), a model-agnostic framework that applies to many modalities and supports early and late fusion scenarios.

Diffusion models have emerged as a powerful class of generative models across various modalities, including image, video, and audio synthesis. However, their deployment is often limited by significant inference latency, primarily due to the inherently sequential nature of the denoising process. While existing parallelization strategies attempt to accelerate inference by distributing computation across multiple devices, they typically incur high communication overhead, hindering deployment on commercial hardware. To address this challenge, we propose $\textbf{ParaStep}$, a novel parallelization method based on a reuse-then-predict mechanism that parallelizes diffusion inference by exploiting similarity between adjacent denoising steps. Unlike prior approaches that rely on layer-wise or stage-wise communication, ParaStep employs lightweight, step-wise communication, substantially reducing overhead. ParaStep achieves end-to-end speedups of up to $\textbf{3.88}$$\times$ on SVD, $\textbf{2.43}$$\times$ on CogVideoX-2b, and $\textbf{6.56}$$\times$ on AudioLDM2-large, while maintaining generation quality.

This paper investigates the role of explicit thinking process in rule-based reinforcement fine-tuning (RFT) for multi-modal large language models (MLLMs). We first extend \textit{Thinking-RFT} to image classification task, using verifiable rewards for fine-tuning~(FT). Experiments show {Thinking-RFT} significantly outperforms supervised FT and yields a cross-dataset generalization effect. We then rethink and question whether explicit thinking in RFT is always necessary and beneficial. Challenging the convention that explicit thinking is crucial for the success of RFT, we introduce \textit{No-Thinking-RFT}, exploring RFT without thinking by introducing a simple equality accuracy reward. We evaluate No-Thinking-RFT on six diverse tasks across different model sizes and types. Experiment results reveal four key findings: \textbf{(1).} Visual perception tasks do not require thinking during RFT, as No-Thinking-RFT consistently outperforms or matches Thinking-RFT across model sizes and types.

Recently, convolutional filters have been increasingly adopted in sequential recommendation for their ability to capture local sequential patterns. However, most of these models complement convolutional filters with self-attention. This is because convolutional filters alone, generally fixed filters, struggle to capture global interactions necessary for accurate recommendation. We propose \textbf{T}ime-\textbf{V}ariant Convolutional Filters for Sequential \textbf{Rec}ommendation (TV-Rec), a model inspired by graph signal processing, where time-variant graph filters capture position-dependent temporal variations in user sequences. By replacing both fixed kernels and self-attention with time-variant filters, TV-Rec achieves higher expressive power and better captures complex interaction patterns in user behavior. This design not only eliminates the need for self-attention but also reduces computation while accelerating inference. Extensive experiments on six public benchmarks show that TV-Rec outperforms state-of-the-art baselines by an average of 7.49\%.

To address these challenges, we propose \texttt{GSAC} (\textbf{G}eneralizable and \textbf{S}calable \textbf{A}ctor \textbf{C}ritic), a framework that couples causal representation learning with meta actor critic learning to achieve both scalability and domain generalization. Each agent first learns a sparse local causal mask that provably identifies the minimal neighborhood variables influencing its dynamics, yielding exponentially tight approximately compact representations (ACRs) of state and domain factors. These ACRs bound the error of truncating value functions to $\kappa$-hop neighborhoods, enabling efficient learning on graphs. A meta actor critic then trains a shared policy across multiple source domains while conditioning on the compact domain factors; at test time, a few trajectories suffice to estimate the new domain factor and deploy the adapted policy. We establish finite sample guarantees on causal recovery, actor-critic convergence, and adaptation gap, and show that \texttt{GSAC} adapts rapidly and significantly outperforms learning-from-scratch and conventional adaptation baselines.

Layer pruning has emerged as a widely used technique for compressing large language models (LLMs). However, existing layer pruning approaches often incur substantial performance degradation. We identify the majority of this degradation to a single yet previously overlooked issue: \textit{the mismatch of activation magnitudes at the pruning interface}.