Logic gate networks replace arithmetic operations with purely Boolean functions for highly efficient inference, yet scaling their width naively causes kernel redundancy and performance saturation. This work identifies that redundancy is most severe at the first gate level and proposes Dynamic Logic Kernel (DLK) and Early-stage Dynamic Logic Kernel (EDLK), which use input-dependent Boolean routing to specialize kernel groups or first-level gate groups. The result is improved kernel diversity and better accuracy-efficiency trade-offs under width scaling, advancing the practicality of discrete logic-based deep learning.

RelationVGGT tackles 3D spatial relation segmentation: given multi-view images, a reference subject mask, and a text relation (e.g., "resting on"), it segments the target object across unposed views without per-scene optimization. The core idea is to replace expensive volumetric relation search with per-pixel relation prediction by fusing semantic DINOv2 features with geometry-aware Pi3 tokens through a subject-conditioned relation transformer. This matters because it moves holistic scene understanding beyond isolated object recognition toward reasoning about how objects are spatially arranged, a capability critical for embodied agents and spatial QA.

Sparse-view 3D reconstruction from only a handful of images is fundamentally ill-posed: standard methods overfit to individual views and leave unseen regions as holes or floaters. HierGF tackles this with a hierarchical 3D Gaussian Splatting pipeline that separates a coarse geometric backbone from a fine-scale field which inherits only spatial means and colors while re-estimating shape and appearance under enhanced pseudo-supervision. Geometry-consistent densification and confidence-weighted diffusion priors are then used to fill gaps and suppress artifacts in under-sampled regions. The authors report strong gains on MipNeRF360, OmniObject3D, LLFF and DTU, alongside a sub-11-minute per-scene reconstruction time on an A100 GPU.

Video benchmarks are supposed to certify that models can understand temporal sequences, yet most can be gamed without watching the video. This paper audits 115 multiple-choice video benchmarks with a five-level attack pyramid that grants attackers access to answer options, question text, other items, single frames, or shuffled videos. After finding severe shortcut vulnerabilities—35 benchmarks break without any visual input and near-duplicates plague 63 of them—the authors screen half a million items down to 840 hardest survivors and compose them into Video-Index, a reproducible meta-benchmark backed by agent-assisted selection and lockfiles.

As video generation scales to minute-long clips, subjects enter and leave the scene, making static-set consistency metrics inadequate. This paper introduces DynSC-Eval, a framework that tracks each subject over its entire visible lifespan and decomposes consistency into local continuity and global identity preservation via six object-level metrics. The authors further derive rewards from these metrics and use DiffusionNFT to post-train video models, demonstrating substantial consistency gains in an autonomous driving testbed and extending them to 30-second generation through curriculum learning.

This paper studies masked diffusion multimodal language models and shows that answer candidates often stabilize retrospectively long before their positions are permanently committed or the rationale-side canvas is filled. The authors introduce two clocks—token stabilization $T_{\mathrm{stab}}$ and commitment $T_{\mathrm{commit}}$—and measure how much of the rationale region $\mathcal{R}$ remains unwritten at stabilization. They pair this with matched-canvas visual-sensitivity ablations and a coverage-accuracy decomposition to explain why prompting changes affect overall accuracy. The analysis matters because it reveals that observing the final rationale alone is insufficient to understand when or how a diffusion model decides its answer.

This paper attacks a growing bottleneck in text-to-image training: how to audit billion-scale recaptioned corpora before spending millions of GPU-hours on downstream training. It proposes a matched-budget audit framework that fixes a text budget $B=64$ and profiles any supervision distribution $D_{\pi,V_c,C}$ across five axes—prompt-side coverage, image-conditioned faithfulness, and caption-surface health—using controllable basic units (CBUs) as a common claim currency. The work matters because it disentangles corpus quality from model-training confounders and releases both a ~490M-caption corpus and a reusable artifact bundle.

Recent video generative models produce stunning visuals but routinely fail at plausible physical interactions and state transitions. This paper proposes a modular synthetic-data pipeline that couples a structured taxonomy with image-editing-based start/end anchors and a novel State-Guided Sampling (SGS) method to smoothly interpolate between them. Fine-tuning a Wan 2.1 model on this curated dataset improves interaction realism, backed by a hybrid VLM-plus-probe evaluator validated against human judgments.

SAM3-ASH tackles fully automated video tracking annotation without any dataset-specific training. Its Generalized Presence Token (GPT) enables SAM3 to process multiple text prompts simultaneously by virtual batching over a single shared image encoding, collapsing per-prompt cost from $\mathcal{O}(N)$ to $\mathcal{O}(1)$. The Annotation and Segmentation Handler (ASH) splits videos of arbitrary length into overlapping temporal chunks and reconciles identities across boundaries via IoU matching, bounding peak GPU memory below 25 GB. Evaluated on eight benchmarks, the pipeline sets a zero-shot HOTA record on MOTS20 and stays within a narrow gap of trained specialists on tasks where mask-based output aligns with the ground-truth format.

Diffusion models with classifier-free guidance (CFG) compute paired conditional and unconditional activations at each denoising step, yet existing post-training quantizers treat those branches as independent sources. This paper identifies the CFG branch space as a structured coding axis—matched activations are strongly correlated and jointly shaped by the guidance direction—and proposes Guidance-Correlation Branch Transform (GCBT), a lightweight $2\times 2$ orthogonal rotation applied per layer before activation quantization. It admits a closed-form per-layer solution, plugs into existing PTQ pipelines with negligible overhead, and offers a complementary, easy-to-adopt lever for low-bit generation fidelity.

FutureWorlds learns robotic world models by turning model-generated alternative futures into training experiences. The core idea unifies three ingredients: diverse beam search to build informative candidate trajectories, candidate-specific bounded memory to maintain diverging rollout histories, and MemSPO—a GRPO-based post-training algorithm that converts decoded trajectory rewards into group-relative advantages with history-consistent scoring. This addresses the exposure bias and token-level likelihood misalignment that plague autoregressive world models, where teacher-forcing training histories diverge from self-generated inference histories.

Feed-forward 3D reconstruction models like VGGT and $\pi^{3}$ rely on expensive quadratic global attention over long image sequences. VASC proposes a training-free sparse attention substitute that ranks block interactions by both query-key relevance and adjacent-value contrast, then tracks unserved computational demand across layers via a lightweight cross-layer memory. It achieves up to $2.29\times$ faster inference than dense VGGT while improving pose and reconstruction accuracy over prior sparse baselines.

This paper tackles video-to-speech (V2S) synthesis, where the central challenge is a one-to-many mapping: visually identical lip movements can correspond to different phonemes. The proposed Watch Your Speech (WYS) framework injects explicit textual conditioning—ground-truth text during training and lip-reader predictions at inference—into a conditional flow matching model via an attention-based fusion module. The key question is whether training-time textual guidance can outperform inference-only classifier guidance while preserving naturalness.

VIEScore2 addresses the shortcoming that most automatic image evaluators return only a scalar quality score without identifying defective regions. It introduces a unified evaluator that represents an image as a sparse $16\times 16$ text grid and jointly predicts perceptual quality, semantic consistency, and defect cell coordinates in a single autoregressive pass. By framing localization as text generation, it enables directly verifiable post-training rewards and supports heterogeneous supervision across generation and editing tasks. The practical value lies in producing human-readable, spatially grounded feedback from one model without task-specific prediction heads.

The paper solves the problem that weight-space task arithmetic is trapped in one architecture because parameter updates share no common coordinate system. Its core idea is the platonic task vector—a model-agnostic object beneath per-model task vectors—and the Universal Task Descriptor (UTD), an architecture-independent matrix of fine-tuning-induced affinity shifts that can be added or negated directly. A closed-form operator folds into the target's last layer, while a low-rank adapter fits the same objective across all layers. The work matters because it enables label-free, weight-free composition of task knowledge across six heterogeneous vision–language and audio–text families.

End-to-end autonomous driving maps raw sensors directly to control actions, yet its real-world bottleneck lies in the training ecosystem rather than model architecture alone. This survey introduces a Data-Strategy-Platform taxonomy that frames training as an interdependent system: data defines what is learnable, strategies govern how learning aligns with driving objectives, and platforms enable scalability. The authors argue the field must shift from raw data volume to value density, from isolated task-specific learners to universal foundation models, and from static offline training to closed-loop training-testing integration.

EyeTAG tackles temporal jitter in video gaze estimation by making gaze motion an explicit variable rather than an implicit deep feature. It feeds back first-order differences of its own recent predictions as a compact kinematic token, then fuses this with face and eye streams via cross-attention and a causal Transformer decoder. The result is a causal streaming model that cuts mean angular error on Gaze360 by roughly a degree and claims subject-invariant motion via translation-invariant differencing.

Foundation models often rely on spurious correlations, yet generating visual counterfactual explanations is usually slow and entangled. This paper proposes DiDAE, a gradient-free method that wraps a frozen visual foundation model in a conditional diffusion decoder and performs closed-form edits along disentangled dictionary atoms. The approach promises up to 2000× speed-ups over iterative methods while enabling targeted model repair via Counterfactual Knowledge Distillation and unsupervised discovery of shortcut directions.

Sign language machine translation remains far behind spoken-language MT because sign languages exploit simultaneous manual and non-manual channels, spatial grammar, and three-dimensional construction. This survey reviews the technical arc from isolated recognition to end-to-end translation and production, embedding every stage in sign language linguistics and deaf-community ethics. The authors argue that progress must be judged not by benchmark scores alone but by whether deaf users actually accept and understand the output, making the manuscript as much a piece of technology assessment as a call for sociotechnical reform.

UniTrackPLA tackles the problem of unifying instruction-guided navigation and dynamic person tracking—two embodied tasks traditionally handled by separate policies—within a single panoramic vision-language-action framework. Its core idea is to fuse omnidirectional perception with a shared continuous waypoint prediction space, using Panoramic-Aware Encoding to adapt perspective pretrained encoders to 360° observations and World-Action Consistency to adaptively decide when to reuse waypoint chunks or replan. This matters because it removes heading-dependent initialization constraints and offers a step toward general-purpose robots that can follow both linguistic routes and moving humans with one model.