Understanding Generalization and Forgetting in In-Context Continual Learning

In-context learning (ICL) derives its power from enabling Large Language Models to adapt to new tasks via prompt-based reasoning alone, entirely bypassing the need for parameter updates. Existing theories primarily study ICL in single-task settings, while real-world prompts often contain sequences of heterogeneous tasks, leaving a gap in understanding whether Large Language Models implicitly perform continual learning during inference. To bridge this gap, we propose the first theoretical framework for in-context continual learning, modeling how a pretrained Transformer processes multiple sequential tasks within a single prompt through shared attention mechanisms. Focusing on linear and masked linear self-attention, we derive error expressions for model predictions under sequential task prompts and analyze their generalization and forgetting behavior. Our results reveal that standard attention mechanisms inevitably induce intertask interference by uniformly or causally aggregating historical contexts, leading to systematic bias. We further provide a bias-variance-interference decomposition of prediction error, characterizing when historical in-context information yields positive transfer or provable negative transfer. This analysis exposes fundamental limits of attention-based continual inference and offers theoretical explanations for order sensitivity and performance degradation in long prompts.

CSPMNet: Pareto-Efficient Automatic Modulation Classification With Learnable Complex Subband Phase Motion

Automatic modulation classification (AMC) is an essential technique for noncooperative spectrum monitoring and intelligent wireless receivers. However, practical AMC models must identify modulation formats from short and noisy I/Q observations while maintaining low computational and storage overhead. Existing deep-learning approaches often improve recognition accuracy by expanding generic neural backbones, which increases deployment cost and weakens their suitability for resource-constrained receivers. To bridge the gap between recognition performance and model efficiency, this letter proposes a Complex Subband Phase-Motion Network, designated as CSPMNet, for lightweight AMC from raw I/Q samples. Specifically, learnable complex subband filters are introduced to adaptively extract frequency-selective baseband responses while preserving the algebraic coupling between in-phase and quadrature components. Then, an amplitude-preserving phase-motion module captures multi-lag temporal rotation dynamics within each subband, and a lightweight temporal classifier performs efficient sequence aggregation. Rigorous experimental evaluations on public RadioML benchmark datasets demonstrate that CSPMNet achieves highly competitive recognition accuracy while requiring substantially lower model complexity than many existing AMC models.

SEI-SHIELD: Robust Specific Emitter Identification Under Label Noise Via Self-Supervised Filtering and Iterative Rescue

Specific Emitter Identification (SEI) provides physical-layer device authentication for wireless communications and Internet of Things (IoT) systems. While deep learning (DL) has significantly advanced SEI performance, label noise severely degrades system reliability in non-cooperative environments. Label noise originates from channel-induced ambiguities, annotation errors, and deliberate data poisoning by intelligent jammers injecting misleading signals. While recent SEI methods attempt to mitigate label noise, they fundamentally rely on corrupted supervised signals to guide sample selection, inevitably leading to confirmation bias and suboptimal feature spaces. To address this challenge, we propose SEI-SHIELD, a robust SEI framework that integrates self-supervised contrastive pre-training with iterative sample selection. Specifically, SEI-SHIELD employs Momentum Contrast (MoCo) with RF-tailored augmentations to extract intrinsically robust, label-independent representations directly from complex-valued I/Q signals. In addition, K-nearest neighbors (KNN)-based noise filtering identifies corrupted samples through neighborhood label consistency analysis in the learned feature space. Furthermore, an iterative rescue mechanism using prediction confidence and prototype cosine similarity progressively recovers correctly labeled hard samples inadvertently discarded during filtering. Comprehensive experiments on the POWDER and ORACLE datasets demonstrate that SEI-SHIELD achieves state-of-the-art (SOTA) accuracy under various noise rates, substantially outperforming existing noise-robust paradigms, including advanced regularization techniques and sample selection frameworks.

MorphSNN: Adaptive Graph Diffusion and Structural Plasticity for Spiking Neural Networks

Spiking Neural Networks (SNNs) currently face a critical bottleneck: while individual neurons exhibit dynamic biological properties, their macro-scopic architectures remain confined within conventional connectivity patterns that are static and hierarchical. This discrepancy between neuron-level dynamics and network-level fixed connectivity eliminates critical brain-like lateral interactions, limiting adaptability in changing environments. To address this, we propose MorphSNN, a backbone framework inspired by biological non-synaptic diffusion and structural plasticity. Specifically, we introduce a Graph Diffusion (GD)mechanism to facilitate efficient undirected signal propagation, complementing the feedforward hierarchy. Furthermore, it incorporates a Spatio-Temporal Structural Plasticity (STSP) mechanism, endowing the network with the capability for instance-specific, dynamic topological reorganization, thereby overcoming the limitations of fixed topologies. Experiments demonstrate that MorphSNN achieves state-of-the-art accuracy on static and neuromorphic datasets; for instance, it reaches 83.35% accuracy on N-Caltech101 with only 5 timesteps. More importantly, its self-evolving topology functions as an intrinsic distribution fingerprint, enabling superior Out-of- Distribution (OOD) detection without auxiliary training. The code is available at anonymous.4open.science/r/MorphSNN-B0BC.

Scaling Agents via Continual Pre-training

Large language models (LLMs) have evolved into agentic systems capable of autonomous tool use and multi-step reasoning for complex problem-solving. However, post-training approaches building upon general-purpose foundation models consistently underperform in agentic tasks, particularly in open-source implementations. We identify the root cause: the absence of robust agentic foundation models forces models during post-training to simultaneously learn diverse agentic behaviors while aligning them to expert demonstrations, thereby creating fundamental optimization tensions. To this end, we are the first to propose incorporating Agentic Continual Pre-training (Agentic CPT) into the deep research agents training pipeline to build powerful agentic foundational models. Based on this approach, we develop a deep research agent model named AgentFounder. We evaluate our AgentFounder-30B on 10 benchmarks and achieve state-of-the-art performance while retains strong tool-use ability, notably 39.9% on BrowseComp-en, 43.3% on BrowseComp-zh, and 31.5% Pass@1 on HLE.

Towards General Agentic Intelligence via Environment Scaling

Advanced agentic intelligence is a prerequisite for deploying Large Language Models in practical, real-world applications. Diverse real-world APIs demand precise, robust function-calling intelligence, which needs agents to develop these capabilities through interaction in varied environments. The breadth of function-calling competence is closely tied to the diversity of environments in which agents are trained. In this work, we scale up environments as a step towards advancing general agentic intelligence. This gives rise to two central challenges: (i) how to scale environments in a principled manner, and (ii) how to effectively train agentic capabilities from experiences derived through interactions with these environments. To address these, we design a scalable framework that automatically constructs heterogeneous environments that are fully simulated, systematically broadening the space of function-calling scenarios. We further adapt a two-phase agent fine-tuning strategy: first endowing agents with fundamental agentic capabilities, then specializing them for domain-specific contexts. Extensive experiments on agentic benchmarks, tau-bench, tau2-Bench, and ACEBench, demonstrate that our trained model, AgentScaler, significantly enhances the function-calling capability of models.

Factorized Disentangled Representation Learning for Interpretable Radio Frequency Fingerprint

In response to the rapid growth of Internet of Things (IoT) devices and rising security risks, Radio Frequency Fingerprint (RFF) has become key for device identification and authentication. However, various changing factors - beyond the RFF itself - can be entangled from signal transmission to reception, reducing the effectiveness of RFF Identification (RFFI). Existing RFFI methods mainly rely on domain adaptation techniques, which often lack explicit factor representations, resulting in less robustness and limited controllability for downstream tasks. To tackle this problem, we propose a novel Disentangled Representation Learning (DRL) framework that learns explicit and independent representations of multiple factors, including the RFF. Our framework introduces modules for disentanglement, guided by the principles of explicitness, modularity, and compactness. We design two dedicated modules for factor classification and signal reconstruction, each with tailored loss functions that encourage effective disentanglement and enhance support for downstream tasks. Thus, the framework can extract a set of interpretable vectors that explicitly represent corresponding factors. We evaluate our approach on two public benchmark datasets and a self-collected dataset. Our method achieves impressive performance on multiple DRL metrics. We also analyze the effectiveness of our method on downstream RFFI task and conditional signal generation task. All modules of the framework contribute to improved classification accuracy, and enable precise control over conditional generated signals. These results highlight the potential of our DRL framework for interpretable and explicit RFFs.

Global Spatio-Temporal Fusion-based Traffic Prediction Algorithm with Anomaly Aware

Traffic prediction is an indispensable component of urban planning and traffic management. Achieving accurate traffic prediction hinges on the ability to capture the potential spatio-temporal relationships among road sensors. However, the majority of existing works focus on local short-term spatio-temporal correlations, failing to fully consider the interactions of different sensors in the long-term state. In addition, these works do not analyze the influences of anomalous factors, or have insufficient ability to extract personalized features of anomalous factors, which make them ineffectively capture their spatio-temporal influences on traffic prediction. To address the aforementioned issues, We propose a global spatio-temporal fusion-based traffic prediction algorithm that incorporates anomaly awareness. Initially, based on the designed anomaly detection network, we construct an efficient anomalous factors impacting module (AFIM), to evaluate the spatio-temporal impact of unexpected external events on traffic prediction. Furthermore, we propose a multi-scale spatio-temporal feature fusion module (MTSFFL) based on the transformer architecture, to obtain all possible both long and short term correlations among different sensors in a wide-area traffic environment for accurate prediction of traffic flow. Finally, experiments are implemented based on real-scenario public transportation datasets (PEMS04 and PEMS08) to demonstrate that our approach can achieve state-of-the-art performance.

RoboMIND: Benchmark on Multi-embodiment Intelligence Normative Data for Robot Manipulation

Developing robust and general-purpose robotic manipulation policies is a key goal in the field of robotics. To achieve effective generalization, it is essential to construct comprehensive datasets that encompass a large number of demonstration trajectories and diverse tasks. Unlike vision or language data that can be collected from the Internet, robotic datasets require detailed observations and manipulation actions, necessitating significant investment in hardware-software infrastructure and human labor. While existing works have focused on assembling various individual robot datasets, there remains a lack of a unified data collection standard and insufficient diversity in tasks, scenarios, and robot types. In this paper, we introduce RoboMIND (Multi-embodiment Intelligence Normative Data for Robot manipulation), featuring 55k real-world demonstration trajectories across 279 diverse tasks involving 61 different object classes. RoboMIND is collected through human teleoperation and encompasses comprehensive robotic-related information, including multi-view RGB-D images, proprioceptive robot state information, end effector details, and linguistic task descriptions. To ensure dataset consistency and reliability during policy learning, RoboMIND is built on a unified data collection platform and standardized protocol, covering four distinct robotic embodiments. We provide a thorough quantitative and qualitative analysis of RoboMIND across multiple dimensions, offering detailed insights into the diversity of our datasets. In our experiments, we conduct extensive real-world testing with four state-of-the-art imitation learning methods, demonstrating that training with RoboMIND data results in a high manipulation success rate and strong generalization. Our project is at https://x-humanoid-robomind.github.io/.

GazeGaussian: High-Fidelity Gaze Redirection with 3D Gaussian Splatting

Gaze estimation encounters generalization challenges when dealing with out-of-distribution data. To address this problem, recent methods use neural radiance fields (NeRF) to generate augmented data. However, existing methods based on NeRF are computationally expensive and lack facial details. 3D Gaussian Splatting (3DGS) has become the prevailing representation of neural fields. While 3DGS has been extensively examined in head avatars, it faces challenges with accurate gaze control and generalization across different subjects. In this work, we propose GazeGaussian, a high-fidelity gaze redirection method that uses a two-stream 3DGS model to represent the face and eye regions separately. By leveraging the unstructured nature of 3DGS, we develop a novel eye representation for rigid eye rotation based on the target gaze direction. To enhance synthesis generalization across various subjects, we integrate an expression-conditional module to guide the neural renderer. Comprehensive experiments show that GazeGaussian outperforms existing methods in rendering speed, gaze redirection accuracy, and facial synthesis across multiple datasets. We also demonstrate that existing gaze estimation methods can leverage GazeGaussian to improve their generalization performance. The code will be available at: https://ucwxb.github.io/GazeGaussian/.