arXiv subjects

Chen Tang

Publications and source records attributed to Chen Tang.

At least 19 recordsLinked to original sources

arXiv2026

vExpert: Virtualizing Expert Storage for Adaptive Load Balancing in Distributed MoE Inference

Mixture-of-Experts (MoE) models have become prevailing, yet their distributed deployment suffers from severe dynamic load imbalance due to the conflict between static expert placement and inherent routing dynamism. While host DRAM offers capacity to avoid static device storage for load balancing, we identify existing offloading works fundamentally fail in distributed inference due to two overlooked challenges: (1) coupled performance impacts of runtime Host-to-Device transfers due to load dynamism and shared PCIe topology; (2) system-level memory overhead incurred by DRAM usage at scale. In this paper, we propose vExpert that shifts DRAM offloading paradigm from partial residency to full redundancy through virtualized expert storage. We pioneer the first systematic performance model that jointly captures H2D latency, PCIe contention, and load imbalance, building a lightweight allocator to dynamically map physical experts to virtual slots runtime. A novel disaggregated manager is designed to reduce memory overhead in multi-process environments. Evaluated on DeepSeek-V3.1 within an 8-node H100 cluster, vExpert improves load balance ratio by up to 55% and achieves 1.2$\times$ end-to-end speedup.

cs.DC↗

arXiv2026

From Stages to States: Unified Iterative Reasoning for MLLMs

Multimodal Chain-of-Thought (MCoT) substantially enhances the complex reasoning capabilities of Multimodal Large Language Models (MLLMs) by explicitly modeling intermediate reasoning processes. However, most learning-based MCoT methods follow a stage-centric paradigm, where different reasoning behaviors are assigned to predefined stages and optimized separately. This results in fragmented reasoning and a training-inference reasoning-state gap. To address these limitations, we propose ReState, a unified iterative reasoning framework that reformulates multimodal reasoning from a stage-driven pipeline into a state-driven dynamic process. It comprises two complementary components: Unified Reasoning State Modeling (URSM) and Closed-Loop Reasoning-State Learning (CRSL). URSM unifies reasoning construction, correct reasoning preservation, and erroneous reasoning correction as reasoning-state transitions within a shared state space, enabling these heterogeneous reasoning behaviors to be learned jointly. CRSL further incorporates self-generated historical reasoning states during training and recursively updates the current state during inference, thereby mitigating the reasoning-state gap and enabling iterative self-refinement without requiring an additional reflection module, critic, or independent correction stage. Furthermore, we construct VersaCoT-10K, a high-quality CoT dataset spanning eight visual reasoning domains, which provides diverse supervision for multi-domain reasoning and reasoning-state refinement. Extensive experiments demonstrate that ReState exhibits strong adaptability across different MLLM backbones and achieves strong performance on multiple public reasoning benchmarks, supporting the shift from stage-centric reasoning toward state-centric reasoning. Our data and code are available at https://restate-project.github.io

cs.AI↗

arXiv2026

TS-DP: Reinforcement Speculative Decoding For Temporal Adaptive Diffusion Policy Acceleration

Diffusion Policy (DP) excels in embodied control but suffers from high inference latency and computational cost due to multiple iterative denoising steps. Existing diffusion model (DMs) acceleration methods improve inference speed by reducing computational overhead, as DMs for image generation are computation-bound. In this work, we reveal that DP are I/O bound, and thus propose using speculative decoding, a parallel computing framework, to accelerate DP for the first time. However, it is non-trivial to address the following challenges: (1) How to preserve the denoising quality of the base model at reduced computational cost under time-varying task difficulty in embodied settings; and (2) How to dynamically allocate computation according to task difficulty in such environments. Therefore, we propose Temporal-aware Reinforcement-based Speculative Diffusion Policy (TS-DP), the first framework that enables speculative decoding for DP with temporal adaptivity. First, to handle dynamic environments where task difficulty varies over time, we distill a Transformer-based drafter to imitate the base model and replace its costly denoising calls. Second, an RL-based scheduler further adapts to the time-varying task difficulty by adjusting speculative parameters to maintain accuracy while improving efficiency. Extensive experiments show that TS-DP achieves up to a 4.17x NFE reduction and a 3.2x wall-clock speedup with a draft acceptance rate of over 85% in simulation. Real-world experiments further demonstrate a 3.8x NFE reduction while improving success rates on real-time tasks (e.g., a 23% gain on the conveyor-belt task), reaching an inference frequency of 26 Hz and enabling real-time control while maintaining comparable task performance. The code is available at https://github.com/ChildTang/TS-DP.

cs.LG↗

arXiv2026

PhysEvolve: Verifier-grounded Self-evolution for Physical Reasoning with Executable Worlds

Physical reasoning underpins how embodied systems understand and act in the physical world, yet vision-language models (VLMs) remain unreliable on questions about dynamics and physical causality. Progress is critically constrained by the scarcity of grounded counterfactual training data, as observational videos capture only the factual trajectory that occurred. The self-evolving paradigm offers a route beyond fixed datasets by enabling models to construct their own training problems. Extending this paradigm to physics, however, requires an external source of truth that can both validate model-proposed environments and verify alternative physical outcomes under intervention. We introduce \physevolve{}, a verifier-grounded self-evolving framework that trains VLMs for causal physical reasoning from model-proposed executable worlds and simulation-authored tasks. At each round, a shared VLM proposes executable world programs and evidence schemas; a deterministic physics engine freezes each valid world, simulates matched baseline and body-elided counterfactual branches, and synthesizes discrete, answer-bound visual tasks. Exact verifier feedback from task solving updates both the shared policy via GRPO and an adaptive curriculum over physical intents. Starting from Qwen3-VL-8B-Instruct, \physevolve{} achieves targeted gains across PhysBench, CausalPhys, and ContPhy without human intervention data, improving PhysBench Throwing by $8.8$ points and ContPhy Counterfactual Removal and Goal-Pit by $7.4$ and $24.4$ points. These gains concentrate on interaction-sensitive and altered-world questions, demonstrating zero-shot transfer of causal principles to unseen complex dynamics (cloth, fluid, soft bodies) and real-world video.

cs.AI↗

arXiv2026

GUIFlow: Temporal Redundancy-Aware Incremental Inference for GUI Agents

GUI agents built on vision-language models interact with continuously evolving interfaces, yet conventional inference repeatedly recomputes representations of largely overlapping observations and histories. This mismatch makes multi-step interaction costly, while changes in context, interface content, and action arguments prevent straightforward cache reuse. We introduce GUIFlow, a framework for full-pipeline incremental inference in multi-image GUI interaction. Our central insight is that advancing a decision need not require rebuilding all supporting representations: each interaction step can instead selectively update computational states maintained across the trajectory. To support new decisions as the context evolves, GUIFlow adapts how past computation is reused: historical KV states are position-aligned and injected into the current context; change signals guide selective updates of current-observation states in the vision encoder and language model; and earlier action responses serve as speculative drafts verified by the target model. Together, these mechanisms enable cross-step reuse throughout visual encoding, prefill, and decoding without additional training, a separate draft model, or changes to the model-visible inputs and action interface. Experiments demonstrate up to 3.39x geometric-mean speedup across offline multi-image benchmarks and up to 3.49x in online, closed-loop interaction on OSWorld, while maintaining task performance broadly comparable to the dense baseline.

cs.CV↗

arXiv2026

CoLink: Global Compression and Knowledge Linking for Multi-Turn Dialogue Modeling

Extracting salient information from long-horizon dialogue histories remains a significant challenge for Large Language Models (LLMs) due to the restricted context windows and efficient retrieval challenges. Current methods typically rely on history organization or local token-level optimization, suffering from information loss, retrieval redundancy, and myopic attention. Inspired by the view in cognitive science that dialogue constitutes sense-making and memory association, we propose CoLink, a framework that enhances LLMs' global context awareness in domain-specific multi-turn dialogues through information linking and compression. CoLink first clusters historical contexts into topic graphs to establish cross-semantic associations with external knowledge. Guided by the adaptive control of thought, it then retrieves and activates knowledge via synergizing semantic relevance and graph-structural topology. Crucially, to achieve effective compression of history and knowledge, we ground multi-turn dialogues in information theory and derive a principled optimization objective for extracting the core sense-making information. Coupled with a competitive slot mechanism, CoLink condenses the global context into dedicated memory and knowledge slots, efficiently distilling lengthy, redundant inputs into an extremely compact latent representation. Extensive empirical results on both public and our MMD datasets validate CoLink's superiority and generalizability in knowledge linking and sense modeling.

cs.CL↗

arXiv2026

A Differentiable Chassis Safety Layer for Intelligent Vehicle Motion Control

Learning-based vehicle motion control requires coordination between diverse task objectives and state-dependent chassis safety constraints. Conventional cascaded architectures may compromise task performance when safety filters modify nominal commands independently of policy learning. This article proposes a differentiable chassis safety layer that modularly integrates task-oriented control with explicit physical constraints. Actuator limits, individual-tire force constraints, and vehicle handling envelopes are incorporated into sequential quadratic programs. Each projected command propagates the predicted vehicle dynamics and updates subsequent constraints, enabling temporal gradient coupling through implicit differentiation. A two-stage learning strategy combines expert imitation with risk-guided supervision and conditional intervention guidance to coordinate nominal control decisions with chassis safety constraints. Closed-loop evaluations under unseen scenarios, emergency maneuvers, and ablation studies demonstrate that the proposed method improves safety-performance coordination and exploits vehicle dynamic potential while maintaining handling stability.

cs.RO↗

arXiv2026

TerraTransfer++: Demonstration-Free End-to-End Planner

End-to-end autonomous driving has achieved state-of-the-art performance on benchmarks and real-world deployments. Its standard training recipe, however, is expensive across all existing paradigms: imitation learning requires fleet-scale driving logs whose behavior distribution only thinly covers the safety-critical tail; closed-loop reinforcement learning in a photorealistic simulator, while offering diverse behaviors, is bottlenecked by sample-efficiency, per-step cost of rendering, and a forward pass through a large vision backbone. Self-play in vectorized simulators changes the economics: at millions of rollout steps per second, one can obtain a state distribution with natural diversity in collisions, near-misses, and recoveries that no driving log contains, and a highly reactive policy emerges from a basic driving-quality reward. Our approach exploits this asymmetry. We pretrain a single policy by self-play, distill its rollouts into a flow-matching trajectory teacher that also plans, and train lightweight adapters on a frozen pretrained vision backbone to reproduce the latent space of either teacher through the action KL divergence, a batch-relational low-rank structural loss, and, for the planner, a trajectory-matching loss. Every target traces back to the self-play policy, so alignment never supervises against a logged trajectory: a paired dataset of (image, scene-state) frames suffices, with no need for the curated expert demonstrations that imitation pretraining is built on. From this one self-play teacher we obtain two end-to-end students: a closed-loop policy that outperforms every published imitation-based end-to-end method on the photorealistic 3D Gaussian splatting closed-loop benchmark HUGSim, and a planner that reaches 59.4 EPDMS on NAVSIM navhard, above every published method including DriveZero-Scale (57.1), and 95.5 PDMS on navtest, without human demonstrations.

cs.RO↗

arXiv2026

ADT: All-Atom Diffusion Tokenizer for 3D Molecular Structures

Discrete structural tokenization provides a promising route to autoregressive 3D molecular generation. Yet, existing tokenizers face a trade-off between compact representation and all-atom reconstruction: atom-level tokens preserve atomic details but produce long sequences, while compact residue-level protein tokens typically omit side-chain geometry. To overcome this limitation, we propose All-atom Diffusion Tokenizer (ADT), which decouples token granularity from reconstruction resolution. ADT uses one structural token per atom for small molecules and one token per protein residue that jointly represents backbone and side-chain geometry. To mitigate geometric distortion introduced by quantization, we leverage the generative capacity of diffusion models to recover fine-grained atomic geometry beyond deterministic decoding from discrete tokens. Extensive experiments across small molecules and proteins demonstrate that ADT achieves strong reconstruction from compact discrete representations while enabling high-quality autoregressive generation of both small molecules and all-atom protein structures. These results establish compact structural tokens as an effective discrete interface for all-atom molecular reconstruction and generation.

cs.LG↗

arXiv2026

Mem4Act: Learning Effective Latent Memory for Long-Horizon Robotic Manipulation

Long-horizon robotic manipulation fundamentally requires memory: agents must retain task-relevant information from past observations and interactions and use it to guide future actions. Existing approaches typically preserve raw historical observations, compressed visual tokens, or language summaries, but these representations either expose the action policy to increasingly long histories or risk discarding task-relevant temporal, semantic, and perceptual information. We introduce Mem4Act, a framework for learning effective latent memory as a fixed-size representation explicitly optimized for long-horizon robotic control. Mem4Act addresses three coupled challenges: how to efficiently cover long histories, what information memory should retain, and how to ensure the policy actually relies on memory. To this end, it organizes history via multi-granularity temporal aggregation, preserves complementary semantic and perceptual information through question-answering supervision and key-frame DINO feature reconstruction, and employs memory-first two-stage training to encourage greater use of memory by the action policy. The resulting memory captures both high-level task state and fine-grained perceptual evidence while providing a bounded interface to the action policy. Across RoboMME, RMBench, and four real-world manipulation tasks, Mem4Act achieves state-of-the-art performance on memory-dependent control, with success rates of 49.88%, 61.0%, and 81.5%, respectively. Extensive ablations further demonstrate the importance of temporal aggregation, complementary memory supervision, and memory-first training.

cs.RO↗

arXiv2026

PROTEAN: Beyond De Novo Generation - Can LLM Agents Perform Text-Guided Protein Structure Editing?

Generative AI has achieved remarkable success in de novo protein design, yet protein engineering typically begins with existing, characterized proteins. The primary demand is targeted local editing, e.g., engineering a specific region to optimize affinity, alter immunogenicity, or improve stability, while strictly preserving core catalytic motifs and overall scaffold integrity. However, no standardized framework currently exists for carrying out and evaluating such edits. In current generative-design workflows, researchers often need to manually specify complex low-level parameters and rely on repetitive, trial-and-error sampling and evaluation across generative backends to search for viable conformations. While large language models (LLMs) offer an intuitive interface to express these goals using natural language, prevailing structure-generation models cannot directly interpret high-level human instructions. To bridge this gap, we introduce PROTEAN, a standardized benchmark for text-guided protein structural editing. PROTEAN comprises 380 sub-tasks spanning six intervention families and evaluates a two-stage pipeline in which an LLM translates natural-language goals into backend-specific geometric constraints and a generative model remodels the protein backbone. A two-tier evaluation separates geometric compliance, including target attainment and off-target scope fidelity, from biophysical realizability assessed by inverse folding and refolding consistency. Across five frontier LLMs paired with three backends, strict end-to-end biophysical success starts at just 7.0%. Ablation studies show that LLM constraint compilation is already highly accurate, whereas the main bottlenecks are backend-specific scope leakage and refolding failure. PROTEAN provides an open testbed for developing autonomous, multi-backend macromolecular design agents.

cs.CE↗

arXiv2026

BRIDGE: Selective Boundary Repair with Dependency Graph Evidence in Multi-Agent Systems

Multi-agent collaboration has emerged as a critical paradigm for tackling complex, long-horizon tasks; however, while coordination protocols continue to advance, systemic failures at the inter-agent collaboration interface remain a pervasive bottleneck. In collaborative workflows, systems frequently suffer from \emph{broken handoffs}---where critical task context is omitted, misrouted, or ignored despite each agent emitting locally plausible responses. To address this challenge, we present \method, a verifiable runtime control framework that treats handoff repair as selective intervention over explicit dependency graphs. Central to \method is an integrated fault-attribution and calibration pipeline, \dataset, which decouples execution dependencies and injects counterfactual handoff perturbations to provide empirical supervision for runtime intervention. Leveraging this diagnostic foundation, \method dynamically compiles observable interaction traces into a provenance graph to pinpoint unresolved dependencies, without accessing private agent states. It then evaluates candidate repairs through an empirical gate that weighs expected gains against disruption risks, and executes repairs only when formal local postconditions hold. Extensive evaluations across demanding benchmarks demonstrate that \method boosts task success by up to 12.3 absolute percentage points on SILO-BENCH and 11.1 on TeamBench. Across both backbones, \method maintains strict cost-effectiveness, adding as little as 2.0\% marginal token overhead. Furthermore, on 562 handoff-relevant traces from the MAST benchmark, \method achieves 69.8\% F1 in a single pass, establishing a reliable operational boundary for verifiable coordination repair.

cs.CL↗

arXiv2026

H2Table: Hierarchical Hypergraph-Enhanced Large Language Models for Complex Table Reasoning

Tables are ubiquitous across diverse domains, yet reasoning over them remains a significant challenge for modern large language models (LLMs). Current approaches typically linearize tables into sequences, inherently overlooking their intrinsic two-dimensional and hierarchical structure. To address this, we propose H2Table (Hierarchical Hypergraph-Enhanced Table Reasoning), a novel framework that represents complex tables as hierarchical nested hypergraphs. To process this representation, we design a tailored hypergraph encoder to facilitate message passing between hyperedges (headers) and nodes (cells), thereby perceiving the semantic entailment relationships between them within complex tables. Furthermore, we introduce a set of learnable query vectors acting as a lightweight bridge to extract representative structural embeddings from the encoder into the LLM. Experimental results demonstrate that our approach effectively handles complex table question answering tasks with hierarchical nested headers. Notably, on the HiTab dataset, H2Table achieves an average improvement of 22.88% over state-of-the-art baselines on highly complex tables with a nesting depth of four. Our code is available at: https://github.com/lila120/h2table.

cs.AI↗

arXiv2026

RapidMoE: Exploiting Cross-Asymmetry via Adaptive Residual Offloading for Large-Scale MoE Inference

The widespread adoption of Mixture-of-Experts (MoE) has created a growing need for deployment on heterogeneous platforms. However, it exposes a fundamental mismatch between the algorithmic demands of large-scale MoE and the disparate characteristics of hardware.Existing CPU-GPU hybrid inference systems fail to resolve this as they either encounter PCIe bandwidth bottlenecks when loading experts to GPUs, or rely heavily on CPU computation. Consequently, this leads to low resource utilization and inevitable violations of fixed latency budgets as parameters scale. In this paper, we identify and exploit Cross-Asymmetry--a structural alignment between the algorithmic workload skew of MoE routing and the physical disparity of heterogeneous hardware. To this end, we introduce RapidMoE, a residual offloading system for efficient large-scale MoE inference. We propose how RapidMoE leverages a residual-split framework to enable offloading paradigm shift from expert-level to bit-level, which unfolds across three key dimensions: (1) data representation, enabling compact and decoupled storage; (2) routing strategy, partitioning computation into dual paths aligned with hardware capabilities; (3) execution parallelism, scheduling a balanced storage-compute workload across devices. We further employ a novel Unified Multi-Level Importance Arbitration to adaptively adjust the critical expert set at runtime, ensuring the accuracy-latency Pareto frontier. These innovations exploit inherent cross-asymmetry, fundamentally breaking the algorithm-hardware misalignment. Experimental results show that RapidMoE achieves up to 3.5x speedup in decoding and 2.1x speedup in prefill compared to state-of-the-art (SOTA) offloading systems.

cs.DC↗

arXiv2026

The More It Says, the More You Pay: A Black-Box Audit of Provider-Side Token Inflation in LLM Services

In pay-per-token LLM services, the more a model says, the more users pay. Dishonest providers can covertly manipulate generation to inflate output tokens while largely preserving task utility. We define such manipulation as a Provider-Side Token Inflation Attack (PTIA) and instantiate five representative attacks at the query, prompt, representation, and model levels of the provider-controlled pipeline. Our experiments show that each attack increases mean output length to more than 10.2x the clean baseline, demonstrating PTIA's financial appeal and feasibility at multiple stages of generation. Yet auditing PTIA from black-box responses is difficult for users. Our key observation is PTIA saturation: an initial attack sharply lengthens output, but further strengthening or composition has much less effect. We trace this saturation to stopping behavior: an initial PTIA sharply lowers the end-of-sequence token probability, whereas further intervention lowers it only marginally. Building on this insight, we design a lightweight single-probe audit that applies a controlled lengthening intervention. Under PTIA, the probe induces far fewer additional tokens than under normal service. The audit requires neither a trusted local reference model nor historical clean responses, and its separately issued original and probed requests resemble ordinary traffic, making evasion difficult. Across four open-weight models, it achieves an average detection rate of 85.1% with false-positive rates below 2%. Across 15 real LLM API services, the audit flags 7 for PTIA-consistent behavior.

cs.CR↗

arXiv2026

SCAPE: Scenario-Conditioned Simulation-Augmented Policy Evaluation

Reliable performance evaluation is a central bottleneck for deploying robot-learning policies in real-world conditions. Real-world testing is faithful but costly and difficult to scale, whereas simulation-based testing scales easily but is inevitably biased by the sim-to-real gap. Existing simulation-augmented methods combine limited real-world rollouts with abundant simulation proxies, but focus on performance averaged over initial conditions and deployment settings. Such population-level averages obscure scenario-specific variation and provide limited guidance about when and where a policy can be safely deployed. We propose SCAPE, a scenario-conditioned simulation-augmented policy evaluation framework that predicts scenario-conditioned real-world policy performance using limited paired sim-and-real samples and large-scale simulation rollouts. SCAPE corrects sim-to-real bias in simulation labels before training the prediction model and calibrates prediction uncertainty through conformal prediction. We validate SCAPE on autonomous driving and quadruped velocity tracking. In sim-to-sim studies, SCAPE reduces scenario-level prediction error by 4.9%/34.7% (driving) and 14.5%/27.7% (quadruped) relative to scene-conditioned neural and aggregate statistical baselines on average. We further evaluate a velocity-tracking policy deployed on a physical Unitree Go2. SCAPE also improves testing sample efficiency, produces narrower calibrated prediction intervals, generalizes better to out-of-distribution scenarios, and enables fine-grained deployment strategies.

cs.RO↗

arXiv2026

KA2L: A Knowledge-Aware Active Learning Framework for LLMs

Fine-tuning large language models (LLMs) with high-quality knowledge has been shown to enhance their performance effectively. However, there is a paucity of research on the depth of domain-specific knowledge comprehension by LLMs and the application of targeted active learning to improve their expertise. To address this gap, we introduce the Knowledge-Aware Active Learning (KA2L) framework. This framework assesses LLMs' mastery of specific knowledge points to aid in constructing unanswerable or unknowable questions through latent space analysis. This active learning strategy enhances training efficiency by focusing on knowledge the model has yet to master, thereby minimizing redundancy in learning already acquired information. This study innovatively employs a knowledge distribution probing technique to examine the hidden states of specific Transformer layers and identify the distribution of known and unknown knowledge within the LLM. Additionally, a hidden-state decoding method is proposed to generate numerous unknown questions in natural language from the latent knowledge space. In our experiments, we selected nine open-source LLMs to validate the effectiveness of the proposed framework. Results indicate that KA2L not only significantly reduces 50% annotation and computation costs across two open-domain and one vertical-domain dataset but also achieves better performance, offering valuable insights into active learning strategies for LLMs. The code is available at https://github.com/greenjerry/KA2L.

cs.CL↗

arXiv2026

DAPD: Dual-Anchored Policy Distillation

On-policy (self) distillation (OPSD) is increasingly adopted for language-model post-training. It strengthens the teacher with privileged information but can induce a privilege illusion: the student learns privilege-dependent behavior it cannot reproduce from its inference-time context, yet behaves as if the training-time privileged information remained available, ultimately degrading performance. In this paper, we identify information asymmetry between the privileged teacher and the student at inference as the root cause of this failure in OPSD. To resolve this asymmetry, we propose Dual-Anchored Policy Distillation (DAPD), a unified framework with two levels of anchoring. Dual-Path Anchoring (DPA) introduces a self-conditioned bridge and aligns reference and rollout behavior along two matched-information paths, preventing privilege-dependent behavior from being transferred to the inference-time student. Dual-Source Anchoring (DSA) applies these paths in both reference-to-rollout and rollout-to-reference directions, reducing reliance on privileged reference guidance while preserving correctness supervision. Extensive experiments show that DAPD significantly alleviates privilege illusion, outperforming OPSD on Qwen3-4B by +2.00 points on average across tasks. Notably, its gains persist across scales, reaching +2.69 at 4B and +2.78 at 32B.

cs.AI↗