Based on the message-passing paradigm, there has been an amount of research proposing diverse and impressive feature propagation mechanisms to improve the performance of GNNs. However, less focus has been put on feature transformation, another major operation of the message-passing framework. In this paper, we first empirically investigate the performance of the feature transformation operation in several typical GNNs. Unexpectedly, we notice that GNNs do not completely free up the power of the inherent feature transformation operation. By this observation, we propose the Bi-directional Knowledge Transfer (BiKT), a plug-and-play approach to unleash the potential of the feature transformation operations without modifying the original architecture. Taking the feature transformation operation as a derived representation learning model that shares parameters with the original GNN, the direct prediction by this model provides a topological-agnostic knowledge feedback that can further instruct the learning of GNN and the feature transformations therein. On this basis, BiKT not only allows us to acquire knowledge from both the GNN and its derived model but promotes each other by injecting the knowledge into the other. In addition, a theoretical analysis is further provided to demonstrate that BiKT improves the generalization bound of the GNNs from the perspective of domain adaption. An extensive group of experiments on up to 7 datasets with 5 typical GNNs demonstrates that BiKT brings up to 0.5% - 4% performance gain over the original GNN, which means a boosted GNN is obtained. Meanwhile, the derived model also shows a powerful performance to compete with or even surpass the original GNN, enabling us to flexibly apply it independently to some other specific downstream tasks.

We introduce Heterogeneous Agent Collaborative Reinforcement Learning (HACRL), a new Reinforcement Learning from Verifiable Reward (RLVR) problem that addresses the inefficiencies of isolated multi-agent on-policy optimization. HACRL enables collaborative optimization with independent execution: heterogeneous agents share verified rollouts during training to mutually improve, while operating independently at inference time. Unlike LLM-based multi-agent reinforcement learning (MARL), HACRL does not require coordinated deployment, and unlike on-/off-policy distillation, it enables bidirectional mutual learning among heterogeneous agents rather than one-directional homogeneous teacher-to-student transfer. Building on this problem, we propose HACPO, a collaborative RL algorithm that enables principled rollout sharing to maximize sample utilization and cross-agent knowledge transfer. To mitigate capability discrepancies and policy distribution shifts, HACPO introduces four tailored mechanisms with theoretical guarantees on unbiased advantage estimation. Extensive experiments across diverse heterogeneous model combinations and reasoning benchmarks show that HACPO consistently improves all participating agents, outperforming GSPO with double rollouts by an average of 3.6

Reinforcement Learning from Human Feedback (RLHF) is a pivotal technique for aligning large language models (LLMs) with human preferences, yet it is susceptible to reward overoptimization, in which policy models overfit to the reward model, exploit spurious reward patterns instead of faithfully capturing human intent. Prior mitigations primarily relies on surface semantic information and fails to efficiently address the misalignment between the reward model (RM) and the policy model caused by continuous policy distribution shifts. This inevitably leads to an increasing reward discrepancy, exacerbating reward overoptimization. To address these limitations, we introduce R2M (Real-Time Aligned Reward Model), a novel lightweight RLHF framework. R2M goes beyond vanilla reward models that solely depend on the semantic representations of a pretrained LLM. Instead, it leverages the evolving hidden states of the policy (namely policy feedback) to align with the real-time distribution shift of the policy during the RL process. This work points to a promising new direction for improving the performance of reward models through real-time utilization of feedback from policy models.

The honesty of Large Language Models (LLMs) is increasingly important for safe deployment in high-stakes domains. However, this crucial trait is severely undermined by supervised fine-tuning (SFT), a common technique for model specialization. Existing recovery methods rely on data-intensive global parameter adjustments, implicitly assuming that SFT deeply corrupts the models' ability to recognize their knowledge boundaries. However, we observe that fine‑tuned LLMs still preserve this ability; what is damaged is their capacity to faithfully express that awareness. Building on this, we propose Honesty-Critical Neurons Restoration (HCNR) to surgically repair this suppressed capacity. HCNR identifies and restores key expression-governing neurons to their pre-trained state while harmonizing them with task-oriented neurons via Hessian-guided compensation. Experiments on four QA tasks and five LLM families demonstrate that HCNR effectively recovers 33.25% of the compromised honesty while achieving at least 2.23x speedup with over 10x less data compared to baseline methods, offering a practical solution for trustworthy LLM deployment.

Graph Foundation Models (GFMs) are pre-trained on diverse source domains and adapted to unseen targets, enabling broad generalization for graph machine learning. Despite that GFMs have attracted considerable attention recently, their vulnerability to backdoor attacks remains largely underexplored. A compromised GFM can introduce backdoor behaviors into downstream applications, posing serious security risks. However, launching backdoor attacks against GFMs is non-trivial due to three key challenges. (1) Effectiveness: Attackers lack knowledge of the downstream task during pre-training, complicating the assurance that triggers reliably induce misclassifications into desired classes. (2) Stealthiness: The variability in node features across domains complicates trigger insertion that remains stealthy. (3) Persistence: Downstream fine-tuning may erase backdoor behaviors by updating model parameters. To address these challenges, we propose GFM-BA, a novel Backdoor Attack model against Graph Foundation Models. Specifically, we first design a label-free trigger association module that links the trigger to a set of prototype embeddings, eliminating the need for knowledge about downstream tasks to perform backdoor injection. Then, we introduce a node-adaptive trigger generator, dynamically producing node-specific triggers, reducing the risk of trigger detection while reliably activating the backdoor. Lastly, we develop a persistent backdoor anchoring module that firmly anchors the backdoor to fine-tuning-insensitive parameters, enhancing the persistence of the backdoor under downstream adaptation. Extensive experiments demonstrate the effectiveness, stealthiness, and persistence of GFM-BA.

On-policy distillation (OPD) aligns a student with a teacher on trajectories sampled from the student itself, reducing the train-test state mismatch of offline distillation. The same feedback loop can nevertheless be unstable: each update changes both the policy and the states on which the next update is computed. We introduce WDL-OPD, a mixture-constrained co-training method with two trainable policies. An anchor policy generates every rollout, an auxiliary policy evaluates the same visited states, and a geometric mixture of their token distributions is matched to a frozen teacher by reverse KL. Both policies receive gradient. We show that freezing the auxiliary recovers an anchor-plus-contrast proxy target closely related to OPD^2 and W2S-OPD, whereas joint training creates branch-level degrees of freedom that a static delta cannot express. In recorded Qwen3 experiments at 1.7B and 4B scale, WDL-OPD produces the strongest student checkpoint in each of four scale-domain settings. It raises MATH500 accuracy from 0.630 to 0.685 at 4B and from 0.521 to 0.585 at 1.7B. In code generation, seven single-policy OPD configurations exhibit entropy growth or trajectory degradation, while co-training reaches independently re-evaluated development scores of 0.637 and 0.375. Because several comparisons differ in curriculum or initialization, these results support a stabilization hypothesis rather than a universal causal claim. We provide the exact training algorithm, failure evidence, and the controlled comparison matrix needed to test that hypothesis.

Table Question Answering (TQA), which aims to answer natural language questions over tabular data, has recently attracted growing interest in the database community. While state-of-the-art (SOTA) methods based on online Large Language Models (LLMs) achieve remarkable accuracy, they suffer from several drawbacks, including data privacy risks, high latency, high cost, and overthinking due to excessively long reasoning chains. To address these challenges, we propose SPARQ (Sufficient Precise Adaptive Routing for TableQA), a cost-efficient TQA framework designed for robust offline deployment. SPARQ extends the operator pool for table reasoning and introduces an adaptive query routing mechanism that dynamically selects optimal operators, assisted by a verifier with rollback/fallback strategies. Through extensive evaluations, we demonstrate that SPARQ achieves remarkable performance in an offline setting: it improves accuracy by over 5% on WikiTQ and Tab Fact datasets, while reducing the average end-to-end latency by up to 10.19× on consumer-grade hardware (e.g., RTX 4090). To the best of our knowledge, this is the first systematic framework that deploys offline LLMs to achieve SOTA performance for TQA under realistic hardware constraints, balancing both effectiveness and efficiency. Code, full version and artifacts are provided. 11https://github.com/authurlord/SPARQ

Dynamic graphs are ubiquitous in real-world systems, and building generalizable dynamic Graph Foundation Models has become a frontier in graph learning. However, dynamic graphs from different domains pose fundamental challenges to unified modeling, as their semantic and temporal patterns are inherently inconsistent, making the multi-domain pre-training difficult. Consequently, the widely used "pretrain-then-finetune" paradigm often suffers from severe negative knowledge transfer. To the best of our knowledge, there exists no multi-domain dynamic GFM. In this work, we propose DyGFM, a Dynamic Graph Foundation Model over multiple domains based on decoupled and divergence-conditioned prompting. To disentangle transferable semantics from the domain-specific dynamics, we introduce a dual-branch pre-training strategy with semantic-temporal decoupling. To alleviate negative transfer during domain adaptation, we further develop a cross-domain routing mechanism with divergence-aware expert selection. To enable efficient downstream fine-tuning, we design a divergence-conditioned prompt generator that injects lightweight, learnable graph prompts tailored to semantic and temporal traits. Extensive experiments on continuous dynamic graph benchmarks demonstrate that DyGFM consistently outperforms 12 state-of-the-art baselines on both node classification and link prediction tasks, achieving superior effectiveness and efficiency.

Re-ranking plays a crucial role in modern information search systems by refining the ranking of initial search results to better satisfy user information needs. However, existing methods show two notable limitations in improving user search satisfaction: inadequate modeling of multifaceted user intents and neglect of rich side information such as visual perception signals. To address these challenges, we propose the Rich-Media Re-Ranker framework, which aims to enhance user search satisfaction through multi-dimensional and fine-grained modeling. Our approach begins with a Query Planner that analyzes the sequence of query refinements within a session to capture genuine search intents, decomposing the query into clear and complementary sub-queries to enable broader coverage of users' potential intents. Subsequently, moving beyond primary text content, we integrate richer side information of candidate results, including signals modeling visual content generated by the VLM-based evaluator. These comprehensive signals are then processed alongside carefully designed re-ranking principle that considers multiple facets, including content relevance and quality, information gain, information novelty, and the visual presentation of cover images. Then, the LLM-based re-ranker performs the holistic evaluation based on these principles and integrated signals. To enhance the scenario adaptability of the VLM-based evaluator and the LLM-based re-ranker, we further enhance their capabilities through multi-task reinforcement learning. Extensive experiments demonstrate that our method significantly outperforms state-of-the-art baselines. Notably, the proposed framework has been deployed in a large-scale industrial search system, yielding substantial improvements in online user engagement rates and satisfaction metrics.

Diffusion Transformers (DiTs) achieve strong video generation quality but suffer from high inference cost due to dense 3D attention, motivating sparse attention techniques for improving efficiency. However, existing training-free sparse attention methods for video generation still face two unresolved limitations: ignoring layer heterogeneity in attention pruning and ignoring query-key coupling in block partitioning, which hinder a better quality-speedup trade-off. In this work, we uncover a critical insight: attention sparsity is an intrinsic layer-wise property, with only minor variation across different inputs. Motivated by this observation, we propose SVOO, a training-free sparse attention framework for fast video generation via offline layer-wise sparsity profiling and online bidirectional co-clustering. Specifically, SVOO adopts a two-stage paradigm: (i) offline layer-wise sensitivity profiling to derive intrinsic per-layer pruning levels, and (ii) online block-wise sparse attention via a bidirectional co-clustering algorithm. Extensive experiments on seven widely used video generation models demonstrate that SVOO achieves a superior quality-speedup trade-off over state-of-the-art methods, delivering up to 1.93x speedup while maintaining a PSNR of up to 29 dB on Wan2.1. Code is available at: https://github.com/Mutual-Luo/SVOO.

Reinforcement learning has become a central post-training paradigm for improving LLM and agent capabilities. Yet existing RL post-training methods share a common blind spot: they construct local learning signals from sampled trajectories, rewards, or feedback-conditioned targets, then update the policy without explicitly verifying whether the resulting policy outperforms its predecessor. Optimizing these local signals does not necessarily produce a better policy, while finite sampling, generation stochasticity and feedback noise can further widen this gap. We argue that the missing ingredient is policy improvement feedback: the ability to measure progress across policy iterations. We introduce Policy Improvement Reinforcement Learning (PIRL), which formulates inter-iteration performance gain as an explicit objective structurally aligned with final task performance. Building on PIRL, we propose Policy Improvement Policy Optimization (PIPO), a plug-in closed-loop framework that verifies the previous update against a sliding-window historical performance anchor. PIPO uses this improvement feedback to modulate the local learning signal of the base policy optimization algorithm, reinforcing updates associated with measured progress and suppressing those associated with performance drops. We provide theoretical evidence that PIPO locally aligns policy updates with the PIRL improvement objective. Experiments on mathematical reasoning, code, tool-use, and self-distillation settings show that PIPO yields consistent gains across PPO, group-relative, and self-distillation policy optimization families.

Foundation models have emerged as a dominant paradigm in machine learning, enabling broad generalization and efficient adaptation across diverse tasks and domains. While this paradigm has achieved remarkable success in language and vision data, its extension to structured data remains far less understood. Foundation models for structured data are an emerging yet highly impactful research area with a rapidly growing body of literature. In this survey, we provide a systematic analysis of foundation models for structured data, focusing on tabular, time series, and graph data, covering over 150 representative methods. We analyze the intrinsic properties and inductive biases of structured data, clarify the core concepts of foundation models, and conduct an in-depth analysis of the key challenges that hinder the development of foundation models for structured data. Building on these insights, we organize existing approaches into a coherent taxonomy based on tokenization, architectures, pre-training objectives, and adaptation strategies. Finally, we discusse merging research directions and open problems, aiming to provide guidance toward more principled and scalable foundation models for structured data.

Multi-domain graph pre-training has emerged as a pivotal technique in developing graph foundation models. While it greatly improves the generalization of graph neural networks, its privacy risks under membership inference attacks (MIAs), which aim to identify whether a specific instance was used in training (member), remain largely unexplored. However, effectively conducting MIAs against multi-domain graph pre-trained models is a significant challenge due to: (i) Enhanced Generalization Capability: Multi-domain pre-training reduces the overfitting characteristics commonly exploited by MIAs. (ii) Unrepresentative Shadow Datasets: Diverse training graphs hinder the obtaining of reliable shadow graphs. (iii) Weakened Membership Signals: Embedding-based outputs offer less informative cues than logits for MIAs. To tackle these challenges, we propose MGP-MIA, a novel framework for Membership Inference Attacks against Multi-domain Graph Pre-trained models. Specifically, we first propose a membership signal amplification mechanism that amplifies the overfitting characteristics of target models via machine unlearning. We then design an incremental shadow model construction mechanism that builds a reliable shadow model with limited shadow graphs via incremental learning. Finally, we introduce a similarity-based inference mechanism that identifies members based on their similarity to positive and negative samples. Extensive experiments demonstrate the effectiveness of our proposed MGP-MIA and reveal the privacy risks of multi-domain graph pre-training.

Recent studies have shown that Graph Neural Networks (GNNs) exhibit significant bias in class-imbalanced tasks, i.e., they usually perform well on majority-class with abundant training samples but struggle with minority-class nodes. Existing approaches tackle this issue by designing specialized GNNs or training strategies tailored to minority-class nodes. Though effective, these approaches mainly or even exclusively focus on minority-class nodes, leading to a degraded performance on majority-class nodes that GNNs typically handle well. To address this, we propose IMMix, a novel framework based on mixup augmentation strategy to improve the generalization of GNNs under class imbalance. To achieve this, IMMix first uses trainable prototypes to create a well-structured embedding hypersphere. Within this space, it identifies critical pairs consisting of a minority node and an auxiliary node from another class near the decision boundaries. IMMix then introduces SelectiveMixup, which synthesizes new nodes based on the selected node pairs while preserving class-specific information. This process helps capture the intrinsic distribution discrepancies between minority-class nodes and other classes, thereby enhancing the model’s generalization to both majority and minority nodes. Comprehensive experiments on real-world datasets, spanning both long-tailed and natural class-imbalanced settings, demonstrate that IMMix enhances GNNs’ performance across diverse backbones and consistently outperforms state-of-the-art baselines.

Recent advancements in large reasoning models (LRMs) have greatly improved their capabilities on complex reasoning tasks through Long Chains of Thought (CoTs). However, this approach often results in substantial redundancy, impairing computational efficiency and causing significant delays in real-time applications. Recent studies show that longer reasoning chains are frequently uncorrelated with correctness and can even be detrimental to accuracy. In a further in-depth analysis of this phenomenon, we surprisingly uncover and empirically verify that LRMs implicitly know the appropriate time to stop thinking, while this capability is obscured by current sampling paradigms. Motivated by this, we introduce SAGE (Self-Aware Guided Efficient Reasoning), a novel sampling paradigm that unleashes this efficient reasoning potential. Furthermore, integrating SAGE as mixed sampling into group-based reinforcement learning (SAGE-RL) enables SAGE-RL to effectively incorporate SAGE-discovered efficient reasoning patterns into standard pass@1 inference, markedly enhancing both the reasoning accuracy and efficiency of LRMs across multiple challenging mathematical benchmarks.

Exploration is not an optional behavior in natural intelligence; it is an evolutionary principle underlying the emergence and adaptation of intelligence. Curiosity, play, and deliberate probing emerge as evolved responses to uncertainty, enabling organisms to construct internal models, expand competence, and preserve adaptability in changing environments. We argue that this evolutionary logic is equally indispensable for artificial general intelligence (AGI): exploration is not a heuristic appended to learning, but the mechanism through which generality becomes possible. We develop a unified view of Epistemic Exploration for agentic systems: the capacity of an agent to actively acquire information that reduces uncertainty about the world, seek experiences at the boundary of its current capabilities and convert them into durable capability improvement, and preserve epistemic reachability as the readiness and ability to adapt when the world changes. This view yields three criteria: Information Gain, Value Improvement, and Epistemic Reachability. We then introduce an exploration-centered five-level trajectory toward AGI, in which each level is characterised by a distinct exploration capacity and exploration serves as the transition mechanism among levels: • Responder: minimal or no explicit epistemic exploration; the system mainly relies on learned input-output mappings and local token-level variation. • Reasoner: reasoning-space exploration enables hypothesis search, deliberate reasoning trajectories, branching, backtracking, and self-verification beyond reactive response generation. • Agent: interaction-space exploration extends internal deliberation into embodied perception, tool use, memory, and closed-loop action under partial observability. • Prospector: imagination-space exploration uses learned world models to simulate counterfactual futures, reduce the cost and risk of real interaction, and support long-horizon policy improvement. • Ecosystem: coordination-space exploration enables collectives of heterogeneous agents to co-evolve roles, shared representations, communications, and collaborative strategies beyond the limits of any single agent. Finally, we conclude with evaluation principles and open challenges for building exploration-centric agents that continually reduce uncertainty, improve their own capabilities, and maintain readiness to adapt beyond predefined tasks.

Reinforcement Learning from Verifier Rewards (RLVR) has emerged as a widely used approach for post-training large language models on reasoning tasks, with group-based methods such as GRPO and its variants gaining broad adoption. These methods rely on group-relative advantage estimation to avoid learned critics, yet its theoretical properties remain poorly understood. In this work, we uncover a fundamental issue of group-based RL: the group-relative advantage estimator is inherently biased relative to the true (expected) advantage. We provide the first theoretical analysis showing that it systematically underestimates advantages for hard prompts and overestimates them for easy prompts, leading to imbalanced exploration and exploitation. To address this issue, we propose History-Aware Adaptive Difficulty Weighting (HA-DW), an adaptive reweighting scheme that adjusts advantage estimates based on an evolving difficulty anchor and training dynamics. Both theoretical analysis and experiments on five mathematical reasoning benchmarks demonstrate that HA-DW consistently improves performance when integrated into GRPO and its variants. Our results suggest that correcting biased advantage estimation is critical for robust and efficient RLVR training.

Diffusion Language Models (DLMs) have recently demonstrated remarkable capabilities in natural language processing tasks. However, the potential of Retrieval-Augmented Generation (RAG), which shows great successes for enhancing large language models (LLMs), has not been well explored, due to the fundamental difference between LLM and DLM decoding. To fill this critical gap, we systematically test the performance of DLMs within the RAG framework. Our findings reveal that DLMs coupled with RAG show promising potentials with stronger dependency on contextual information, but suffer from limited generation precision. We identify a key underlying issue: Response Semantic Drift (RSD), where the generated answer progressively deviates from the query's original semantics, leading to low precision content. We trace this problem to the denoising strategies in DLMs, which fail to maintain semantic alignment with the query throughout the iterative denoising process. To address this, we propose Semantic-Preserving REtrieval-Augmented Diffusion (SPREAD), a novel framework that introduces a query-relevance-guided denoising strategy. By actively guiding the denoising trajectory, SPREAD ensures the generation remains anchored to the query's semantics and effectively suppresses drift. Experimental results demonstrate that SPREAD significantly enhances the precision and effectively mitigates RSD of generated answers within the RAG framework.