Add to Chrome
Add to Firefox
Add to EdgeVisMMOE: Exploiting Visual-Expert Affinity for Efficient Visual-Language MoE Offloading
May 07, 2026
Large-scale vision-language mixture-of-experts (VL-MoE) models provide strong multimodal capability, but efficient deployment on memory-constrained platforms remains difficult. Existing MoE offloading systems are largely designed for text-centric workloads and become much less effective for visual-heavy inputs, where large numbers of visual tokens induce broader and less predictable expert accesses. We present VisMMoE, a VL-MoE offloading system built on a single systems insight: pruning redundant visual tokens can improve offloading not only by reducing computation, but also by reshaping expert demand. We refer to this effect as \textit{visual-expert affinity}: token pruning makes expert accesses more concentrated within layers and more stable across layers, producing a smaller and more predictable expert working set. Guided by this insight, VisMMoE combines affinity-aware token compression, lookahead expert prediction, and cache/pipeline orchestration to improve expert locality and prefetch effectiveness under tight memory budgets. We implement VisMMoE on multiple frameworks and evaluate it on representative VL-MoE models and benchmarks. VisMMoE improves end-to-end inference performance by up to 2.68x and 1.61x, respectively, over strong baselines for today's VL-MoE deployments while maintaining competitive accuracy.
X-OmniClaw Technical Report: A Unified Mobile Agent for Multimodal Understanding and Interaction
May 07, 2026Inspired by the development of OpenClaw, there is a growing demand for mobile-based personal agents capable of handling complex and intuitive interactions. In this technical report, we introduce X-OmniClaw, a unified mobile agent designed for multimodal understanding and interaction in the Android ecosystem. This unified architecture of perception, memory, and action enables the agent to handle complex mobile tasks with high contextual awareness. Specifically, Omni Perception provides a unified multimodal ingress pipeline that integrates UI states, real-world visual contexts, and speech inputs, leveraging a temporal alignment module to decompose raw data into structured multimodal intent representations. Omni Memory leverages multimodal memory optimization to enhance personalized intelligence by integrating runtime working memory for task continuity with long-term personal memory distilled from local data, enabling highly context-aware and personalized interactions. Finally, Omni Action employs a hybrid grounding strategy that combines structural XML metadata with visual perception for robust interaction. Through Behavior Cloning and Trajectory Replay, the system captures user navigation as reusable skills, enabling precise direct-access execution. Demonstrations across diverse scenarios show that X-OmniClaw effectively enhances interaction efficiency and task reliability, providing a practical architectural blueprint for the next generation of mobile-native personal assistants.
A cross-modal network for facial expression recognition
May 06, 2026Deep neural networks enriched with structural information have been widely employed for facial expression recognition tasks. However, these methods often depend on hierarchical information rather than face property to finish expression recognition. In this paper, we propose a cross-modal network with strong biological and structural information for facial expression recognition (CMNet). CMNet can respectively learn expression information via face symmetry on a whole face, left and right half faces to extract complementary facial features. To prevent negative effect of biological and structural information fusion, a salient facial information refinement module can obtain salient facial expression information to improve stability of an obtained facial expression classifier. To reduce reliance on unilateral facial features, a half-face alignment optimization mechanism is designed to align obtained expression information of learned left and right half faces. Our experimental results demonstrate that CMNet outperforms several novel methods, i.e., SCN and LAENet-SA for facial expression recognition. Codes can be obtained at https://github.com/hellloxiaotian/CMNet.
FSFM: A Biologically-Inspired Framework for Selective Forgetting of Agent Memory
Apr 22, 2026For LLM agents, memory management critically impacts efficiency, quality, and security. While much research focuses on retention, selective forgetting--inspired by human cognitive processes (hippocampal indexing/consolidation theory and Ebbinghaus forgetting curve)--remains underexplored. We argue that in resource-constrained environments, a well-designed forgetting mechanism is as crucial as remembering, delivering benefits across three dimensions: (1) efficiency via intelligent memory pruning, (2) quality by dynamically updating outdated preferences and context, and (3) security through active forgetting of malicious inputs, sensitive data, and privacy-compromising content. Our framework establishes a taxonomy of forgetting mechanisms: passive decay-based, active deletion-based, safety-triggered, and adaptive reinforcement-based. Building on advances in LLM agent architectures and vector databases, we present detailed specifications, implementation strategies, and empirical validation from controlled experiments. Results show significant improvements: access efficiency (+8.49%), content quality (+29.2% signal-to-noise ratio), and security performance (100% elimination of security risks). Our work bridges cognitive neuroscience and AI systems, offering practical solutions for real-world deployment while addressing ethical and regulatory compliance. The paper concludes with challenges and future directions, establishing selective forgetting as a fundamental capability for next-generation LLM agents operating in real-world, resource-constrained scenarios. Our contributions align with AI-native memory systems and responsible AI development.
Ternary Memristive Logic: Hardware for Reasoning Realized via Domain Algebra
Apr 20, 2026Memristive crossbars store numerical weights needing aggregation and decoding; a single junction means nothing alone. This paper presents a fundamentally different use: each junction stores a complete, domain-scoped logical assertion (holds/negated/undefined). Ternary resistance states encode these values directly. We establish a structure-preserving mapping from a domain algebra to crossbar topology: domains become isolated arrays, specialization becomes directed wiring, relation typing controls inheritance gates, and cross-domain links become explicit registers. The physical layout thus embodies the algebra; changing wiring changes reasoning semantics. We detail an ICD-11 respiratory disease classification chip (1,247 entities, ~136k 1T1R junctions) enabling domain scoping, three-valued logic, transitive cascade, typed inheritance, and cross-axis queries. Behavioral simulation (sigma_log=0.15, SNR=20dB) shows error-free operation across 100,000 trials per task with wide tolerance margins. Where prior work unified representation and computation in software, this work unifies them in hardware: reading one junction answers one question, without symbolic interpretation.
Neural Dynamic GI: Random-Access Neural Compression for Temporal Lightmaps in Dynamic Lighting Environments
Apr 14, 2026High-quality global illumination (GI) in real-time rendering is commonly achieved using precomputed lighting techniques, with lightmap as the standard choice. To support GI for static objects in dynamic lighting environments, multiple lightmaps at different lighting conditions need to be precomputed, which incurs substantial storage and memory overhead. To overcome this limitation, we propose Neural Dynamic GI (NDGI), a novel compression technique specifically designed for temporal lightmap sets. Our method utilizes multi-dimensional feature maps and lightweight neural networks to integrate the temporal information instead of storing multiple sets explicitly, which significantly reduces the storage size of lightmaps. Additionally, we introduce a block compression (BC) simulation strategy during the training process, which enables BC compression on the final generated feature maps and further improves the compression ratio. To enable efficient real-time decompression, we also integrate a virtual texturing (VT) system with our neural representation. Compared with prior methods, our approach achieves high-quality dynamic GI while maintaining remarkably low storage and memory requirements, with only modest real-time decompression overhead. To facilitate further research in this direction, we will release our temporal lightmap dataset precomputed in multiple scenes featuring diverse temporal variations.
Chain-of-Models Pre-Training: Rethinking Training Acceleration of Vision Foundation Models
Apr 14, 2026In this paper, we present Chain-of-Models Pre-Training (CoM-PT), a novel performance-lossless training acceleration method for vision foundation models (VFMs). This approach fundamentally differs from existing acceleration methods in its core motivation: rather than optimizing each model individually, CoM-PT is designed to accelerate the training pipeline at the model family level, scaling efficiently as the model family expands. Specifically, CoM-PT establishes a pre-training sequence for the model family, arranged in ascending order of model size, called model chain. In this chain, only the smallest model undergoes standard individual pre-training, while the other models are efficiently trained through sequential inverse knowledge transfer from their smaller predecessors by jointly reusing the knowledge in the parameter space and the feature space. As a result, CoM-PT enables all models to achieve performance that is mostly superior to standard individual training while significantly reducing training cost, and this is extensively validated across 45 datasets spanning zero-shot and fine-tuning tasks. Notably, its efficient scaling property yields a remarkable phenomenon: training more models even results in higher efficiency. For instance, when pre-training on CC3M: i) given ViT-L as the largest model, progressively prepending smaller models to the model chain reduces computational complexity by up to 72%; ii) within a fixed model size range, as the VFM family scales across 3, 4, and 7 models, the acceleration ratio of CoM-PT exhibits a striking leap: from 4.13X to 5.68X and 7.09X. Since CoM-PT is naturally agnostic to specific pre-training paradigms, we open-source the code to spur further extensions in more computationally intensive scenarios, such as large language model pre-training.
Reasoning as Data: Representation-Computation Unity and Its Implementation in a Domain-Algebraic Inference Engine
Apr 13, 2026Every existing knowledge system separates storage from computation. We show this separation is unnecessary and eliminate it. In a standard triple is_a(Apple, Company), domain context lives in the query or the programmer's mind. In a CDC four-tuple is_a(Apple, Company, @Business), domain becomes a structural field embedded in predicate arity. Any system respecting arity automatically performs domain-scoped inference without external rules. We call this representation-computation unity (RCU). From the four-tuple structure, three inference mechanisms emerge: domain-scoped closure, typed inheritance, and write-time falsification via cycle detection per domain fiber. We establish RCU formally via four theorems. RCU is implementable. We present a working symbolic engine (2400 lines Python+Prolog) resolving four engineering issues: rule-data separation, shared-fiber handling, read-only meta-layer design, and intersective convergence. A central result: CDC domain-constrained inference is distinct from Prolog with a domain argument. Two case studies validate the engine. ICD-11 classification (1247 entities, 3 axes) shows fibers resolve multiple inheritance. CBT clinical reasoning shows generalization to temporal reasoning with session turn as ordered domain index. Multi-constraint queries realize CSP arc-consistency with complexity O(m (N/K)^2), confirming the domain lattice's sparsity governs performance. When domain is structural, data computes itself.
Bit-by-Bit: Progressive QAT Strategy with Outlier Channel Splitting for Stable Low-Bit LLMs
Apr 09, 2026Training LLMs at ultra-low precision remains a formidable challenge. Direct low-bit QAT often suffers from convergence instability and substantial training costs, exacerbated by quantization noise from heavy-tailed outlier channels and error accumulation across layers. To address these issues, we present Bit-by-Bit, a progressive QAT framework with outlier channel splitting. Our approach integrates three key components: (1) block-wise progressive training that reduces precision stage by stage, ensuring stable initialization for low-bit optimization; (2) nested structure of integer quantization grids to enable a "train once, deploy any precision" paradigm, allowing a single model to support multiple bit-widths without retraining; (3) rounding-aware outlier channel splitting, which mitigates quantization error while acting as an identity transform that preserves the quantized outputs. Furthermore, we follow microscaling groups with E4M3 scales, capturing dynamic activation ranges in alignment with OCP/NVIDIA standards. To address the lack of efficient 2-bit kernels, we developed custom operators for both W2A2 and W2A16 configurations, achieving up to 11$\times$ speedup over BF16. Under W2A2 settings, Bit-by-Bit significantly outperforms baselines like BitDistiller and EfficientQAT on both Llama2/3, achieving a loss of only 2.25 WikiText2 PPL compared to full-precision models.
Physics-Aligned Spectral Mamba: Decoupling Semantics and Dynamics for Few-Shot Hyperspectral Target Detection
Apr 07, 2026Meta-learning facilitates few-shot hyperspectral target detection (HTD), but adapting deep backbones remains challenging. Full-parameter fine-tuning is inefficient and prone to overfitting, and existing methods largely ignore the frequency-domain structure and spectral band continuity of hyperspectral data, limiting spectral adaptation and cross-domain generalization.To address these challenges, we propose SpecMamba, a parameter-efficient and frequency-aware framework that decouples stable semantic representation from agile spectral adaptation. Specifically, we introduce a Discrete Cosine Transform Mamba Adapter (DCTMA) on top of frozen Transformer representations. By projecting spectral features into the frequency domain via DCT and leveraging Mamba's linear-complexity state-space recursion, DCTMA explicitly captures global spectral dependencies and band continuity while avoiding the redundancy of full fine-tuning. Furthermore, to address prototype drift caused by limited sample sizes, we design a Prior-Guided Tri-Encoder (PGTE) that allows laboratory spectral priors to guide the optimization of the learnable adapter without disrupting the stable semantic feature space. Finally, a Self-Supervised Pseudo-Label Mapping (SSPLM) strategy is developed for test-time adaptation, enabling efficient decision boundary refinement through uncertainty-aware sampling and dual-path consistency constraints. Extensive experiments on multiple public datasets demonstrate that SpecMamba consistently outperforms state-of-the-art methods in detection accuracy and cross-domain generalization.