BrowseComp-$V^3$: A Visual, Vertical, and Verifiable Benchmark for Multimodal Browsing Agents

Multimodal large language models (MLLMs), equipped with increasingly advanced planning and tool-use capabilities, are evolving into autonomous agents capable of performing multimodal web browsing and deep search in open-world environments. However, existing benchmarks for multimodal browsing remain limited in task complexity, evidence accessibility, and evaluation granularity, hindering comprehensive and reproducible assessments of deep search capabilities. To address these limitations, we introduce BrowseComp-$V^3$, a novel benchmark consisting of 300 carefully curated and challenging questions spanning diverse domains. The benchmark emphasizes deep, multi-level, and cross-modal multi-hop reasoning, where critical evidence is interleaved across textual and visual modalities within and across web pages. All supporting evidence is strictly required to be publicly searchable, ensuring fairness and reproducibility. Beyond final-answer accuracy, we incorporate an expert-validated, subgoal-driven process evaluation mechanism that enables fine-grained analysis of intermediate reasoning behaviors and systematic characterization of capability boundaries. In addition, we propose OmniSeeker, a unified multimodal browsing agent framework integrating diverse web search and visual perception tools. Comprehensive experiments demonstrate that even state-of-the-art models achieve only 36% accuracy on our benchmark, revealing critical bottlenecks in multimodal information integration and fine-grained perception. Our results highlight a fundamental gap between current model capabilities and robust multimodal deep search in real-world settings.

Unsafer in Many Turns: Benchmarking and Defending Multi-Turn Safety Risks in Tool-Using Agents

LLM-based agents are becoming increasingly capable, yet their safety lags behind. This creates a gap between what agents can do and should do. This gap widens as agents engage in multi-turn interactions and employ diverse tools, introducing new risks overlooked by existing benchmarks. To systematically scale safety testing into multi-turn, tool-realistic settings, we propose a principled taxonomy that transforms single-turn harmful tasks into multi-turn attack sequences. Using this taxonomy, we construct MT-AgentRisk (Multi-Turn Agent Risk Benchmark), the first benchmark to evaluate multi-turn tool-using agent safety. Our experiments reveal substantial safety degradation: the Attack Success Rate (ASR) increases by 16% on average across open and closed models in multi-turn settings. To close this gap, we propose ToolShield, a training-free, tool-agnostic, self-exploration defense: when encountering a new tool, the agent autonomously generates test cases, executes them to observe downstream effects, and distills safety experiences for deployment. Experiments show that ToolShield effectively reduces ASR by 30% on average in multi-turn interactions. Our code is available at https://github.com/CHATS-lab/ToolShield.

OneVision-Encoder: Codec-Aligned Sparsity as a Foundational Principle for Multimodal Intelligence

Hypothesis. Artificial general intelligence is, at its core, a compression problem. Effective compression demands resonance: deep learning scales best when its architecture aligns with the fundamental structure of the data. These are the fundamental principles. Yet, modern vision architectures have strayed from these truths: visual signals are highly redundant, while discriminative information, the surprise, is sparse. Current models process dense pixel grids uniformly, wasting vast compute on static background rather than focusing on the predictive residuals that define motion and meaning. We argue that to solve visual understanding, we must align our architectures with the information-theoretic principles of video, i.e., Codecs. Method. OneVision-Encoder encodes video by compressing predictive visual structure into semantic meaning. By adopting Codec Patchification, OV-Encoder abandons uniform computation to focus exclusively on the 3.1%-25% of regions rich in signal entropy. To unify spatial and temporal reasoning under irregular token layouts, OneVision-Encoder employs a shared 3D RoPE and is trained with a large-scale cluster discrimination objective over more than one million semantic concepts, jointly capturing object permanence and motion dynamics. Evidence. The results validate our core hypothesis: efficiency and accuracy are not a trade-off; they are positively correlated. When integrated into LLM, it consistently outperforms strong vision backbones such as Qwen3-ViT and SigLIP2 across 16 image, video, and document understanding benchmarks, despite using substantially fewer visual tokens and pretraining data. Notably, on video understanding tasks, OV-Encoder achieves an average improvement of 4.1% over Qwen3-ViT. Codec-aligned, patch-level sparsity is a foundational principle, enabling OV-Encoder as a scalable engine for next-generation visual generalists.

FlowConsist: Make Your Flow Consistent with Real Trajectory

Fast flow models accelerate the iterative sampling process by learning to directly predict ODE path integrals, enabling one-step or few-step generation. However, we argue that current fast-flow training paradigms suffer from two fundamental issues. First, conditional velocities constructed from randomly paired noise-data samples introduce systematic trajectory drift, preventing models from following a consistent ODE path. Second, the model's approximation errors accumulate over time steps, leading to severe deviations across long time intervals. To address these issues, we propose FlowConsist, a training framework designed to enforce trajectory consistency in fast flows. We propose a principled alternative that replaces conditional velocities with the marginal velocities predicted by the model itself, aligning optimization with the true trajectory. To further address error accumulation over time steps, we introduce a trajectory rectification strategy that aligns the marginal distributions of generated and real samples at every time step along the trajectory. Our method establishes a new state-of-the-art on ImageNet 256$\times$256, achieving an FID of 1.52 with only 1 sampling step.

ROMAN: Reward-Orchestrated Multi-Head Attention Network for Autonomous Driving System Testing

Automated Driving System (ADS) acts as the brain of autonomous vehicles, responsible for their safety and efficiency. Safe deployment requires thorough testing in diverse real-world scenarios and compliance with traffic laws like speed limits, signal obedience, and right-of-way rules. Violations like running red lights or speeding pose severe safety risks. However, current testing approaches face significant challenges: limited ability to generate complex and high-risk law-breaking scenarios, and failing to account for complex interactions involving multiple vehicles and critical situations. To address these challenges, we propose ROMAN, a novel scenario generation approach for ADS testing that combines a multi-head attention network with a traffic law weighting mechanism. ROMAN is designed to generate high-risk violation scenarios to enable more thorough and targeted ADS evaluation. The multi-head attention mechanism models interactions among vehicles, traffic signals, and other factors. The traffic law weighting mechanism implements a workflow that leverages an LLM-based risk weighting module to evaluate violations based on the two dimensions of severity and occurrence. We have evaluated ROMAN by testing the Baidu Apollo ADS within the CARLA simulation platform and conducting extensive experiments to measure its performance. Experimental results demonstrate that ROMAN surpassed state-of-the-art tools ABLE and LawBreaker by achieving 7.91% higher average violation count than ABLE and 55.96% higher than LawBreaker, while also maintaining greater scenario diversity. In addition, only ROMAN successfully generated violation scenarios for every clause of the input traffic laws, enabling it to identify more high-risk violations than existing approaches.

Copyright Detective: A Forensic System to Evidence LLMs Flickering Copyright Leakage Risks

We present Copyright Detective, the first interactive forensic system for detecting, analyzing, and visualizing potential copyright risks in LLM outputs. The system treats copyright infringement versus compliance as an evidence discovery process rather than a static classification task due to the complex nature of copyright law. It integrates multiple detection paradigms, including content recall testing, paraphrase-level similarity analysis, persuasive jailbreak probing, and unlearning verification, within a unified and extensible framework. Through interactive prompting, response collection, and iterative workflows, our system enables systematic auditing of verbatim memorization and paraphrase-level leakage, supporting responsible deployment and transparent evaluation of LLM copyright risks even with black-box access.

Journey to the Centre of Cluster: Harnessing Interior Nodes for A/B Testing under Network Interference

A/B testing on platforms often faces challenges from network interference, where a unit's outcome depends not only on its own treatment but also on the treatments of its network neighbors. To address this, cluster-level randomization has become standard, enabling the use of network-aware estimators. These estimators typically trim the data to retain only a subset of informative units, achieving low bias under suitable conditions but often suffering from high variance. In this paper, we first demonstrate that the interior nodes - units whose neighbors all lie within the same cluster - constitute the vast majority of the post-trimming subpopulation. In light of this, we propose directly averaging over the interior nodes to construct the mean-in-interior (MII) estimator, which circumvents the delicate reweighting required by existing network-aware estimators and substantially reduces variance in classical settings. However, we show that interior nodes are often not representative of the full population, particularly in terms of network-dependent covariates, leading to notable bias. We then augment the MII estimator with a counterfactual predictor trained on the entire network, allowing us to adjust for covariate distribution shifts between the interior nodes and full population. By rearranging the expression, we reveal that our augmented MII estimator embodies an analytical form of the point estimator within prediction-powered inference framework. This insight motivates a semi-supervised lens, wherein interior nodes are treated as labeled data subject to selection bias. Extensive and challenging simulation studies demonstrate the outstanding performance of our augmented MII estimator across various settings.

Dissecting Outlier Dynamics in LLM NVFP4 Pretraining

Training large language models using 4-bit arithmetic enhances throughput and memory efficiency. Yet, the limited dynamic range of FP4 increases sensitivity to outliers. While NVFP4 mitigates quantization error via hierarchical microscaling, a persistent loss gap remains compared to BF16. This study conducts a longitudinal analysis of outlier dynamics across architecture during NVFP4 pretraining, focusing on where they localize, why they occur, and how they evolve temporally. We find that, compared with Softmax Attention (SA), Linear Attention (LA) reduces per-tensor heavy tails but still exhibits persistent block-level spikes under block quantization. Our analysis attributes outliers to specific architectural components: Softmax in SA, gating in LA, and SwiGLU in FFN, with "post-QK" operations exhibiting higher sensitivity to quantization. Notably, outliers evolve from transient spikes early in training to a small set of persistent hot channels (i.e., channels with persistently large magnitudes) in later stages. Based on these findings, we introduce Hot-Channel Patch (HCP), an online compensation mechanism that identifies hot channels and reinjects residuals using hardware-efficient kernels. We then develop CHON, an NVFP4 training recipe integrating HCP with post-QK operation protection. On GLA-1.3B model trained for 60B tokens, CHON reduces the loss gap to BF16 from 0.94% to 0.58% while maintaining downstream accuracy.

InfoTok: Regulating Information Flow for Capacity-Constrained Shared Visual Tokenization in Unified MLLMs

Unified multimodal large language models (MLLMs) integrate image understanding and generation in a single framework, with the visual tokenizer acting as the sole interface that maps visual inputs into tokens for downstream tasks. However, existing shared-token designs are mostly architecture-driven and lack an explicit criterion for what information tokens should preserve to support both understanding and generation. Therefore, we introduce a capacity-constrained perspective, highlighting that in shared-token unified MLLMs the visual tokenizer behaves as a compute-bounded learner, so the token budget should prioritize reusable structure over hard-to-exploit high-entropy variations and redundancy. Motivated by this perspective, we propose InfoTok, an information-regularized visual tokenization mechanism grounded in the Information Bottleneck (IB) principle. InfoTok formulates tokenization as controlling information flow from images to shared tokens to multimodal outputs, yielding a principled trade-off between compression and task relevance via mutual-information regularization. We integrate InfoTok into three representative unified MLLMs without introducing any additional training data. Experiments show consistent improvements on both understanding and generation, supporting information-regularized tokenization as a principled foundation for learning a shared token space in unified MLLMs.

Trust but Verify: Adaptive Conditioning for Reference-Based Diffusion Super-Resolution via Implicit Reference Correlation Modeling

Recent works have explored reference-based super-resolution (RefSR) to mitigate hallucinations in diffusion-based image restoration. A key challenge is that real-world degradations make correspondences between low-quality (LQ) inputs and reference (Ref) images unreliable, requiring adaptive control of reference usage. Existing methods either ignore LQ-Ref correlations or rely on brittle explicit matching, leading to over-reliance on misleading references or under-utilization of valuable cues. To address this, we propose Ada-RefSR, a single-step diffusion framework guided by a "Trust but Verify" principle: reference information is leveraged when reliable and suppressed otherwise. Its core component, Adaptive Implicit Correlation Gating (AICG), employs learnable summary tokens to distill dominant reference patterns and capture implicit correlations with LQ features. Integrated into the attention backbone, AICG provides lightweight, adaptive regulation of reference guidance, serving as a built-in safeguard against erroneous fusion. Experiments on multiple datasets demonstrate that Ada-RefSR achieves a strong balance of fidelity, naturalness, and efficiency, while remaining robust under varying reference alignment.