Register-Transfer Level (RTL) simulators are fundamental to modern IC design workflows, yet like any complex software, they frequently contain critical bugs that compromise simulation accuracy. To systematically stress-test these tools, we introduce GI-RTL-Fuzz, a grammar-informed fuzzing framework that leverages Large Language Models (LLMs) to generate high-quality RTL test cases. Our key innovation is guiding the LLM's generation process using Backus-Naur Form (BNF) grammar coverage information, enabling the exploration of complex and rarely-tested language features. By performing differential testing against different RTL simulators, our framework uncovered a significant number of previously unknown inconsistencies in both the linting process and simulation results of major Verilog simulators.
Gongli Xi, Kun Wang, Zeming Gao, Huahui Yi, Haolang Lu, Qianshan Wei, Ye Tian, Wendong Wang
原文摘要
Large multimodal reasoning models solve challenging visual problems via explicit long-chain inference: they gather visual clues from images and decode clues into textual tokens. Yet this capability also increases hallucinations, where the model generates content that is not supported by the input image or the question. To understand this failure mode, we identify \emph{reasoning drift}: during clue gathering, the model over-focuses on question-irrelevant entities, diluting focus on task-relevant cues and gradually decoupling the reasoning trace from visual grounding. As a consequence, many inference-time localization or intervention methods developed for non-reasoning models fail to pinpoint the true clues in reasoning settings. Motivated by these insights, we introduce ClueRecall, a metric for assessing visual clue retrieval, and present ClueTracer, a training-free, parameter-free, and architecture-agnostic plugin for hallucination suppression. ClueTracer starts from the question and traces how key clues propagate along the model's reasoning pathway (question $\rightarrow$ outputs $\rightarrow$ visual tokens), thereby localizing task-relevant patches while suppressing spurious attention to irrelevant regions. Remarkably, \textbf{without any additional training}, ClueTracer improves all \textbf{reasoning} architectures (including \texttt{R1-OneVision}, \texttt{Ocean-R1}, \texttt{MM-Eureka}, \emph{etc}.) by $\mathbf{1.21\times}$ on reasoning benchmarks
Diankun An, Lu Ma, Yongliang Wang, Zhewen Su, Xin Tian
原文摘要
This paper addresses fine-grained prosody control for large-language-model (LLM) based text-to-speech (TTS) synthesis of Mandarin Chinese, motivated by multimedia content-creation scenarios---film dubbing, audiobook production, and interactive voice interfaces---in which prosodic adjustments must be placed on specific characters rather than applied uniformly to whole utterances. We present a data-driven framework in which fine-grained prosodic markers (pitch, duration, energy, and pause) are automatically extracted from raw audio at scale and embedded into the text stream as out-of-vocabulary Unicode symbols. The LLM of an existing TTS system is then fine-tuned to interpret these markers while its architecture remains unchanged, yielding character-level temporal control that complements existing utterance-level style control. An optional text-LLM front-end can translate natural-language instructions into marker sequences. Experiments on 2.15M Mandarin utterances (5,858 hours) with CosyVoice2 as the backbone demonstrate reliable control of character duration, pitch, pause duration, and global speaking rate, with perceptual quality essentially preserved. A cross-tokenizer analysis further indicates that the ASR-optimized tokenizer of the backbone attenuates prosodic variation relative to reconstruction-oriented alternatives---an observation we report as hypothesis-generating evidence for prosody-sensitive system design.
As Generative AI (GenAI) expands beyond individual workflows and becomes increasingly integrated intocollaborative creative work, its potential to address the fundamental challenge of building common groundamong creators with diverse expertise remains underexplored. Specifically, we lack empirical understandingof how GenAI can mediate collaboration asymmetries among heterogeneous creators. To address this gap,we conducted a multi-method study with narrative writers (N = 10) and visual designers (N = 10) in thecontext of game development. Our findings demonstrate that GenAI goes beyond individual assistance toserve as a collaborative mediator by aligning creative intent, sustaining shared understanding, and mitigatinginterpersonal friction among creators. At the same time, we highlight concerns that GenAI may reducecommunication among creators, undermine mutual trust, and reshape interpersonal dynamics and collaborativeprocesses. Based on these insights, we identify design opportunities for GenAI tools that nurture collaborativesynergy while preserving the creativity and values of human work.
The development of new drugs is a time-consuming and expensive process, with average costs estimated to be up to around $2.5 billion. The first step in this process is the design of the new drug, for which de novo drug design, assisted by artificial intelligence, has blossomed in recent years and has revolutionized the field. In particular, generative artificial intelligence has delivered promising results in drug discovery and development, reducing costs and time to solution. However, classical generative models, such as generative adversarial networks (GANs), can be difficult to train and are prone to mode collapse. Quantum computing may be an avenue to overcome some of these issues, providing compact models with far fewer trainable parameters, thereby enhancing the generalizability of GANs. We propose a new latent style-based quantum Wasserstein GAN (QGAN) for drug design, combining a variational autoencoder that maps molecular structures into a latent space with a hybrid quantum GAN. Our pipeline is benchmarked against classical models of the industry-relevant MOSES suite. Style-based noise reuploading at every rotational gate and a Wasserstein gradient penalty are used to improve trainability and to mitigate mode collapse. Training and large-scale sampling with circuits of up to 15 qubits are performed on noiseless quantum simulators to validate our architecture. We present an inference-only demonstration on an IBM Heron quantum computer using 5 qubits, sampling 2,500 new molecules, to investigate quantum noise effects. Our default QGAN, using 110 trainable generator parameters, reaches statistically comparable overall performance and statistically significant gains on selected drug-relevant metrics relative to both the MOSES classical latent GAN baseline, which contains about 700,000 trainable parameters for the generator, and a classical latent GAN with matched generator capacity.
Janik Bürgermeister, Julius Durmann, Martin Bichler, Mete Şeref Ahunbay, Bary Pradelski, Marco Scarsini
原文摘要
We study equilibrium learning in discretized Bayesian games, focusing on first-price auctions as a testbed for gradient-based learning dynamics under private information. While the symmetric independent private values model admits a unique Bayesian correlated equilibrium (BCE) in the continuous limit, standard convergence guarantees for gradient-based learners do not apply: classical no-regret theory predicts convergence only to Bayesian coarse correlated equilibria (BCCE), which can remain large and far from the competitive equilibrium even under fine discretizations. To bridge this gap, we introduce the notion of Bayesian semicoarse correlated equilibrium (BSCCE), an equilibrium concept that refines BCCE by restricting deviations to probability-preserving transformations consistent with projected gradient dynamics. We provide a linear programming characterization of BSCCE with polynomially many variables and constraints and establish the inclusion hierarchy $BCE \subseteq BSCCE \subseteq BCCE$. Our main finding is that the BSCCE set contracts in many cases as the discretization of the Bayesian game is refined. Even with priors for which BCCE fails to concentrate, the diameter of the BSCCE polytope shrinks toward zero, indicating convergence to the unique continuous-game Bayes-Nash equilibrium. These results provide a principled explanation for the observed robustness of gradient-based learning in discretized Bayesian games, showing that these dynamics select a smaller equilibrium set than standard no-regret theory predicts.
Large Language Models (LLMs) often struggle with temporally grounded reasoning because their static parametric memory is misaligned with the dynamic evolution of real-world knowledge. Although RAG and Parameter-Efficient Fine-Tuning (PEFT) provide practical ways to enhance LLMs, they remain insufficient for temporal reasoning. Existing retrieval methods typically prioritize semantic relevance while overlooking temporal validity, event order, and evolutionary dependencies, resulting in evidence that is relevant in isolation but temporally fragmented. Meanwhile, conventional PEFT methods usually employ fixed and input-agnostic low-rank updates, making it difficult to capture time-dependent entity semantics and relation drift. To address these limitations, we propose Temporal Evidence-Parameter Co-Alignment (TEPCA), a framework for LLM reasoning over TKGs. TEPCA aligns LLM reasoning with temporal knowledge at both the evidence and parameter levels. At the evidence level, it introduces time-constrained Dynamic Relevance-Guided Path (DRGP) pruning to preserve globally coherent temporal paths, together with Mutual-Information (MI)-based pruning to retain query-informative local facts under temporal constraints. At the parameter level, it integrates Dynamic-Gated LoRA (DG-LoRA), a timestamp-conditioned low-rank adaptation mechanism that dynamically modulates the effective adaptation subspace according to temporal context. This design enables LLMs to better capture fine-grained entity evolution and time-dependent relation semantics while preserving parameter efficiency. Experimental results on two TKG benchmarks demonstrate that TEPCA achieves state-of-the-art performance. Additional evaluations of DG-LoRA on eight commonsense QA datasets further verify its general input-adaptive capacity, scalability across LLM backbones, and interpretability.
Shortest-path optimality does not guarantee that a memoryless graph generator preserves a prescribed source-target correspondence. We study whether reordering legal shortest edits can retain this correspondence without changing the pairing, mandatory edits, or sampling budget. Under a common Poisson clock, we establish an exact criterion for endpoint coupling recoverability: the source-blind Markov projection attains the unavoidable clock-error floor if and only if routes to distinct targets do not share graph states. Routes to the same target may merge without sacrificing endpoint fidelity. This characterization leads to an exact finite-state routing formulation and a legality-preserving scheduling procedure with recovery certificates. A marker-free construction demonstrates how shared edge structure can retain source distinctions, while a capacity-constrained family exhibits a sharp separation in paired error between edit orders with identical marginal error. Finally, a target-conditioned aliasing bound connects structural route design to the terminal fidelity of a learned graph-only generator. These results characterize when legal edit ordering can preserve a prescribed coupling under fixed operational costs.
Muyuan Ma, Long Cheng, Lijun Han, Haotian Zhai, Xinpan Meng, Xiuze Xia, Houcheng Li, Zeyu Liu
原文摘要
Exoskeletons have shown great potential for augmenting human physical capabilities, yet generating reliable assistance remains challenging because human motion patterns and interaction requirements vary substantially in practice. To address this problem, a dual-loop assistance strategy is proposed to couple multimodal exoskeleton trajectory generation with variable admittance control, where trajectory variance links the outer-loop generator and the inner-loop controller. The generator synthesizes future exoskeleton trajectories through denoising, primarily using whole-body human motion and exoskeleton proprioception, with task descriptions as auxiliary semantic cues. To reduce cross-modal representation mismatch, a three-stage training pipeline is introduced to progressively optimize the human motion encoder, the exoskeleton proprioception encoder, and the trajectory generator. In the inner loop, multiple future trajectories are sampled from the generator to estimate the trajectory mean and variance. The mean is used as the control reference, while the variance provides an uncertainty measure for prediction reliability. This measure is further incorporated into a stiffness regulation law, enabling compliant assistance with input-to-state stability under bounded lumped interaction input. Experimental results demonstrate that the proposed method significantly reduces muscle activation levels compared with standard baselines, while also decreasing users' metabolic energy expenditure relative to the unassisted condition.
Zhiqi Miao, Marko Haralović, Stanislav Levendeev, Sarvesh Angadi, Jannatul Ferdous, Dan-Cristian Ploesteanu, Jiapan Guo, Estefanía Talavera
原文摘要
Few-shot action recognition (FSAR) requires representations that expose discriminative motion dynamics from only a few labeled videos. Recent trajectory-based methods leverage long-term point tracks for explicit temporal correspondence, but often treat trajectories as carriers of sampled visual features or summarize their motion with compact 2D displacement-based descriptors. We instead view an action as the ordered evolution of 2.5D trajectory states and their local geometric interactions. Based on this view, we introduce GeoTRAM, a geometry-aware trajectory state modeling framework for FSAR. GeoTRAM lifts sparse tracks into 2.5D trajectories, where StateMotion models ordered intra-trajectory state evolution and RelMotion captures state-conditioned local interactions through time-varying relative geometry and motion. Quality-filtered spectral sampling provides more reliable trajectory initialization; complementary semantic, motion, and depth cues are separately refined and preserved for joint spatio-temporal modeling; action-guided reweighting and top-$ρ$ trajectory selection support complementary frame- and trajectory-level matching. Experiments on five FSAR benchmarks demonstrate strong gains on the motion-centric Something-Something benchmarks while maintaining competitive performance on more appearance-driven benchmarks and a favorable accuracy--efficiency trade-off.
Coding agents change running software: they patch a service's code or overwrite its stored state, and then act on their own expectation of how the service will respond afterwards. A wrong expectation may surface only several calls later. Function-level code-execution benchmarks omit persistent service state, and agent benchmarks score the actions an agent takes or the final state it reaches. We introduce CTE-Bench, which measures whether a model can predict how an intervention changes a stateful service's future behavior, without asking it to choose actions. Each scenario gives the model Python service code, the calls and responses observed before the intervention, the intervention itself (a source edit or a state overwrite), and 40 fixed future calls; the model predicts every future response, and predictions are checked by executing the service. Three memory protocols control whether the model sees the correct earlier responses, none of them, or its own earlier predictions. CTE-Bench-Core-v1 contains 255 scenarios over six deterministic Python services, giving 10,200 predictions per model. The main score is effect-step value match (VM): exact response equality on the 2,476 future calls whose response the intervention changes. With correct earlier responses revealed, four API-hosted models (DeepSeek V4-Flash, Kimi K2.5, Qwen3.6-35B-A3B, and Claude Sonnet 4.6) reach 54.3%-61.7% effect-step VM. Hiding those responses lowers effect-step VM to 23.2%-29.0%; conditioning on self-generated predictions gives 24.8%-33.2%, and at most 1.2% of scenarios are predicted exactly end to end. Current models thus track intervention effects mainly when correct feedback is supplied, and their errors compound over a rollout. We release CTE-Bench-Core-v1 with its executable oracle, evaluation scripts, and an evaluation card mapping each claim to its protocol.
Feed-forward geometry models such as VGGT recover camera parameters, depth maps, and point maps from one to hundreds of views in a single forward pass, making them powerful backbones for multi-view 3D perception. However, VGGT's global attention is quadratic in the number of frames and dominates end-to-end runtime on long sequences, which has motivated a range of training-free efficiency methods such as token merging. These methods, however, come with a drop in accuracy that makes them hard to use in place of VGGT. We propose GPA-VGGT, a training-free acceleration method that replaces global attention in early blocks with frame attention augmented by a shared cross-frame K/V set. In the remaining blocks, we merge only keys and values while preserving all queries, restore the attention mass of the tokens each merged key represents, and randomly project the matching metric to 128 dimensions to reduce matching cost. Across diverse multi-view geometry tasks, GPA-VGGT provides training-free acceleration at speedups comparable to prior work while preserving the information required for precise geometric estimation, retaining over 99% of VGGT's pose accuracy under AUC@3.
As the advent of large-scale quantum computing threatens classical encryption, Quasi-Cyclic Moderate Density Parity-Check (QC-MDPC) codes have emerged as a foundational element of Post-Quantum Cryptography (PQC). However, decoding these relatively dense codes using classical Belief Propagation algorithms often leads to failures caused by short graph cycles that stall the decoder. In cryptographic applications, these failures are severe security vulnerabilities that leak secret key information. To address this, this paper proposes a Recurrent Bipartite Graph Neural Network (R-BGNN) as proof of concept. By integrating Gated Recurrent Units (GRUs) into the message-passing framework and employing a multi-topology curriculum training strategy, the R-BGNN actively learns to break limit-cycle oscillations. Evaluated across 20 unseen matrix topologies, the proposed architecture significantly outperforms classical Bit-Flipping and Min-Sum baselines, reducing the Decoding Failure Rate (DFR) by 63.7\% at moderate noise levels ($t=10$) while converging in fewer iterations.
Yuanting Fan, Jun Liu, Bin-Bin Gao, Xiaochen Chen, Yuhuan Lin, Zhewei Dai, Jiawei Zhan, Chengjie Wang
原文摘要
Generating realistic visual defects is crucial for alleviating the scarcity of abnormal samples in anomaly detection, yet existing methods often rely on category-specific few-shot adaptation and struggle to generalize across objects, defect types, and domains. This limitation stems in part from the lack of paired defect editing data and the fine-grained nature of defect appearance, where subtle variations in scale, texture, and morphology must be preserved while maintaining consistency with the target scene. To address this challenge, we introduce UDG, a dataset of 300K normal-abnormal-mask-caption quadruplets curated from diverse defect-related scenarios, and present UniDG, a unified defect generation model. UniDG formulates defect synthesis as Defect-Context Editing: it extracts reference defect context with adaptive cropping, organizes reference and target inputs in a structured diptych format, and fuses multimodal conditions through MM-DiT attention. We further develop a two-stage training strategy: Diversity-SFT learns diverse transferable defect priors from paired editing data, while optional Consistency-RFT can further improve reference adherence, local realism, and defect-category consistency. Without per-category fine-tuning or using MVTec-AD/VisA for training, UniDG outperforms prior few-shot anomaly generation and image insertion/editing baselines in both synthesis quality and downstream single- and multi-class anomaly detection/localization.
Antoine Poirot-Bourdain, Alexis Van Straaten, Emmanuel Bacry, Anne-Sophie Jannot, Jean Feydy
原文摘要
This paper introduces survivalGPU, a high-performance Python and R package designed to accelerate survival analysis computations using graphics processing units (GPUs). While the Cox proportional hazards model remains a cornerstone of survival analysis, it can impose substantial computational burdens when applied to large longitudinal data sets. To address this limitation, survivalGPU implements Cox proportional hazards models and weighted cumulative exposure (WCE) models in Python, using PyTorch and KeOps for efficient back-end computation, with interfaces in both Python and R. Numerical validation shows that survivalGPU reproduces the results of reference implementations within the tolerances expected for the selected floating-point precision, with relative errors below 10 -2 in float32 and below 10 -5 in float64. Benchmark experiments demonstrate substantial speedups. For Cox models, coxphGPU() was 5.6x faster than coxph() from the survival package on 1,000,000 patients, and 14.0x faster based on runtimes extrapolated to 1,000 bootstrap replicates on 100,000 patients. Gains were larger for WCE models: wceGPU() was 920.7x faster than WCE() from the WCE package on 100,000 patients, and 1,321.4x faster based on runtimes extrapolated to 1,000 bootstrap replicates. These results show that survivalGPU makes flexible time-dependent survival modeling and bootstrap-based estimation of confidence intervals feasible for large-scale pharmacoepidemiological applications.
Mid-training (MT) is a training phase in LLM training pipelines that continues pretraining (PT) on smaller datasets of higher value. Taking the PT and MT phases as a whole, training usually starts with only PT data and then transitions completely to MT data. In this work, we view this complete transition from PT to MT data as a special case of a broader class of data mixing schedules that specify how the proportions of PT and MT tokens in each batch change over training. Under fixed total training compute and MT token budgets, we ask: Is this complete transition to MT data optimal for minimizing final MT loss? We show that the answer depends on the learning rate (LR) schedule: one data schedule may outperform another under one LR schedule, but changing only the LR schedule can reverse the outcome. Specifically, when the LR decays throughout continued training, introducing MT data earlier and mixing it with PT data lowers the final MT loss compared with placing all MT data at the end of training, but this result is reversed if we use constant LR. We therefore study how to jointly choose when to introduce MT data and when to decay the LR through ablation studies, and find that it is better to decay the LR only after MT data enters training. Finally, we develop a theoretical model that explains why the optimal data schedule for a fixed MT token budget depends on the LR schedule.
Xiaoxi Kang, Lizhen Qu, Lay-Ki Soon, Adnan Trakic, Terry Yue Zhuo, Patrick Charles Emerton, Genevieve Grant
原文摘要
Large Language Models (LLMs), such as ChatGPT, have drawn a lot of attentions recently in the legal domain due to its emergent ability to tackle a variety of legal tasks. However, it is still unknown if LLMs are able to analyze a legal case and perform reasoning in the same manner as lawyers. Therefore, we constructed a novel corpus consisting of scenarios pertain to Contract Acts Malaysia and Australian Social Act for Dependent Child. ChatGPT is applied to perform analysis on the corpus using the IRAC method, which is a framework widely used by legal professionals for organizing legal analysis. Each scenario in the corpus is annotated with a complete IRAC analysis in a semi-structured format so that both machines and legal professionals are able to interpret and understand the annotations. In addition, we conducted the first empirical assessment of ChatGPT for IRAC analysis in order to understand how well it aligns with the analysis of legal professionals. Our experimental results shed lights on possible future research directions to improve alignments between LLMs and legal experts in terms of legal reasoning.
Continuous visual world simulation requires coupling evolving scene geometry with temporally persistent appearance. We present UniWorld, a unified framework that integrates depth trajectories and reference appearance within a shared video generation model. The central design maintains visual and geometric histories on a common timeline. Noise-aligned overlapping continuation places generated history at the prescribed sampling noise level and commits only newly generated latents. Rolling depth latent state preserves the corresponding geometric controls, incrementally reads newly available geometry, and keeps the conditioning fixed within each sampling window. Together, these mechanisms ensure exact preservation of latent history while bounding the active depth state by the window size. We evaluate continuous generation, paired history interventions, and long-horizon behavior across diverse scene categories. The same generator supports estimated, native, and on-demand simulator depth, while cached and incremental execution paths produce bitwise-identical outputs. Appearance changes and geometric control branches operate through the same interface without modifying committed history. The measured steady-state DiT execution path achieves real-time generation throughput, demonstrating the efficiency of UniWorld for continuous visual world simulation. Project page: https://zephipage.github.io/UniWorld.
We prove a pumping-like lemma for languages accepted by one-register alternating finite-memory automata over an infinite alphabet. Every sufficiently long accepted word has a nonempty factor in a bounded suffix that can be repeated under successive powers of a single alphabet permutation. The proof analyzes accepting runs backwards, counting register values by their state profiles within the remaining suffix and summarizing values outside it. We derive effective semi-linearity of the sets of accepted word lengths and prove non-closure under reversal, concatenation, and Kleene star. We also show that this pumping lemma cannot be extended to alternating automata with two registers.
Chumeng Liang, Linxuan Wang, Xinyu Peng, Huabin Liu, XiaoXu Guo, Yuxin Chen, Ge Liu, Guang Lin 等 10 位作者
原文摘要
Speculative decoding accelerates language model inference by verifying multiple draft tokens in a single target-model pass. Recent parallel drafters have achieved breakthrough performance in frontier production models, but their effectiveness deteriorates as the entropy of target distributions increases due to insufficient draft diversity. To overcome this bottleneck without sacrificing parallelism, we introduce UBTree, a parallel drafter that couples a Unigram proposer with a Bigram selector to construct drafting Trees. The unigram proposer is trained with the standard cross-entropy objective to generate candidate tokens independently for each position, while a lightweight bigram selector predicts transition scores between adjacent candidate pairs. Unlike the proposer, the selector is trained with a renormalized KL objective on high-temperature data. This tree-native training broadens the supervision beyond the greedy path, encouraging plausible alternative branches that improve the chance of accepting additional tokens during tree verification. Across seven standardized benchmarks with Qwen3-4B and Qwen3-8B, UBTree achieves an average speedup of $5.84$--$6.94\times$ over autoregressive decoding and outperforms DARTree in all 28 comparisons. Production-scale evaluation further demonstrates UBTree's advantage over frontier baselines such as DSpark.
Jiahui Li, Jiawei Sun, Zixiang Ren, Ming Liu, Jiamin Shi, Ruiteng Zhao, Zhiyang Liu, Liying Liu 等 10 位作者
原文摘要
Recent perception-free end-to-end (E2E) autonomous driving methods bypass explicit perception outputs by compressing dense image patch tokens into compact scene tokens for downstream trajectory generation and scoring. While these scene tokens form a compact visual bottleneck for the planner, they receive supervision solely from the planning objective, providing limited constraints on the encoded visual information. To address this limitation, we introduce Neural Token Reconstruction (NTR), a representation learning framework to directly constrain the compact scene-token bottleneck in perception-free driving. NTR introduces a self-distillation masked latent reconstruction objective that reconstructs masked patch-level latent features using only compact scene tokens as reconstruction memory. This forces reconstruction gradients to pass exclusively through the scene-token bottleneck, encouraging scene tokens to preserve richer and less redundant visual representations for planning. We further introduce semantic priors derived from foundation-model annotations as a weak semantic interface biasing reconstruction targets toward driving-related structures without introducing explicit perception heads. All auxiliary reconstruction components are removed at inference time, leaving the deployed planner unchanged. NTR achieves state-of-the-art performance on three public autonomous driving benchmarks, including 8.0461 RFS on Waymo E2E and 94.1 PDMS / 90.9 EPDMS on NavSim1&2. The learned scene tokens exhibit lower pairwise redundancy and higher effective rank, indicating that effective bottleneck supervision improves both compact visual representation learning and planning performance.
Controllable world simulation requires more than visually plausible frames: scenes must retain persistent state, coordinated actions, and editable temporal structure. Generative world models offer flexible visual synthesis, but their outputs do not always expose the objects and transitions needed for precise revision. Graphics engines provide executable state and direct control, yet translating language into valid operations requires specialized knowledge of assets, geometry, and interdependent APIs. We present CineU (Cinematize from Unreal), an agentic system for language-driven world orchestration and cinematic choreography within Unreal Engine. From a code-based world modeling perspective, an Unreal project serves as a persistent, inspectable world representation, while agent-generated code specifies operations whose effects can be executed and verified. CineU combines hierarchical orchestration, cascaded asset retrieval and geometry-constrained placement, capability-aligned code synthesis, and multi-stage verification with localized repair. Shared scene and timeline context coordinates entities, actions, camera trajectories, and shot boundaries while supporting dependency-aware revisions. Experiments on scene construction, selective editing, multi-shot cinematography, and component ablations show how specialized execution knowledge and engine feedback improve task completion and targeted recovery. CineU illustrates how executable world state can complement learned visual generation to support controllable and revisable simulated worlds.
Boyu Li, Zhe Lin, Eli Shechtman, Ding Li, Hongbo Fu, Yuqian Zhou
原文摘要
We present StreamStudio, a prototype system that explores a new authoring paradigm for streaming video generation. Instead of entering a prompt and waiting for a complete video before making changes, creators can interact with the generation as it unfolds: steering what comes next, returning to earlier moments, and exploring different continuations. In this way, StreamStudio shifts creative control from post-generation revision to continuous interaction with the generative process. Our contribution centers on this authoring paradigm and the interaction mechanisms that make streaming generation directly manipulable. StreamStudio is designed as a plug-and-play interface that can work with compatible generation and editing models, allowing its interaction model to evolve alongside advances in underlying video models. As streaming models become faster and easier to control in the future, such an interface may support more flexible and controllable video creation, including richer edits and multimodal interventions during generation.
Neural routing solvers rarely stop at their first solution: constructed tours are usually refined by search, such as sampling or local search. Neural improvement methods learn this search as a policy that repeatedly applies local moves, but existing policies degrade sharply beyond their training size. We argue that size generalization depends on whether the information that determines the value of a move reaches the policy in a size-independent form, and identify three places where existing policies fall short: the representation of the incumbent solution, the scale of the input features, and the computation of move gains. We address all three in NICO, a learned 2-opt policy for routing that represents the incumbent by its solution edges, one token per edge without positional encodings, describes them with scale-free features, and adds the exact gain of each move to its logit, so that the network learns a correction to greedy selection. A controlled factorial study shows that edge tokens outperform node tokens with positional encodings, that the exact gains are necessary for transfer in the tested configuration, and that scale-free features help only together with them. Trained on instances with at most 100 nodes, NICO reaches a 1.1% gap on the Euclidean TSP with 1000 nodes, against 10.7% for greedy 2-opt and 36% for the best prior learned policy. It also improves the tours of constructive and diffusion-based solvers, is the best improvement method on road-network costs, and on the CVRP outperforms every classical control.
Conformal Prediction (CP) provides robust uncertainty guarantees for predictive models, but is typically applied post hoc, which misaligns model training with the conformal goal of producing efficient (i.e., narrow) intervals. We propose SPACR (Single-Pass Adaptive Conformal Regressor), a novel method for directly training uncertainty-aware regressors within a differentiable loss. SPACR jointly optimizes accuracy, efficiency, and validity without batch-splitting or a predefined confidence level during training. As a result, a single SPACR model yields valid prediction intervals at multiple confidence levels during inference, avoiding the costly retraining required by methods like Directly Optimized Inductive Conformal Regression (DOICR). Experiments on diverse tabular and image datasets show that SPACR consistently gives tighter intervals and better coverage-efficiency trade-offs compared to standard CP and DOICR, while significantly reducing computational costs relative to retraining-dependent baselines.
Vincent Jolivet, Lucas Terriel, Chahan Vidal-Gor{è}ne
原文摘要
The large-scale digitisation of medieval manuscripts creates a growing need for automated iconographic indexing. We evaluate two complementary paradigms for this task on 9,735 full-folio images annotated with 2,333 TIMEL concepts: generative Vision-Language Models (VLM) and supervised discriminative classifiers (SDC). The VLM experiments evaluate fine-tuning, structured JSON decoding, taxonomybased candidates generation, and deterministic reranking. The discriminative experiments evaluate flat, hierarchy-aware, and patch-attention heads over a frozen DINOv3 encoder. Our results highlight key functional differences: ML-Decoder (SDC) achieves the strongest overall retrieval performance (precision@1 of 0.696, MRR of 0.806), with loss calibration (Asymmetric Loss) serving as the primary performance lever. While SDCs provide superior precision and reliability within a closed vocabulary, VLMs offer strong semantic generalization and broad taxonomic coverage, particularly through structured JSON targets and verify-calibrate reranking (hierarchical F1 of 0.559). Furthermore, model behavior strongly depends on concept abstraction: SDCs excel at perceptually grounded descriptive concepts, whereas VLMs provide useful hypotheses for complex interpretative scenes. Finally, we find that enforcing explicit hierarchy within model architectures limits coverage compared to spatial feature aggregation. Overall, rather than competing approaches, SDCs and VLMs emerge as complementary components for future hybrid indexing systems combining structured retrieval with exploratory annotation assistance. As a methodological contribution, we define the soft F1 metric to provide a taxonomy-aware evaluation that credits semantically close predictions based on lowest common ancestors.
AI agents now install packages, call model APIs with their operators' keys and change code in repositories that deploy to production. The network, credentials and write access they need are also how they can leak data, steal a key or change a system nobody approved. Sandboxes contain an agent but leave little evidence of what crossed their boundary; tracing records what happened but contains nothing. We present Tracekit Box, an open-source system (MIT, v0) that does both. Each job runs in its own Firecracker micro-VM with no network card. The only ways out are doors on the host, served over vsock: commands, outbound connections, secrets, model calls and the workspace patch. At every door a Tracekit signer the agent cannot reach decides what passes and signs it into one run per box. Names are resolved on the host; private and cloud-metadata addresses are always refused; secrets never enter the VM, which sees placeholders the host proxy swaps for real values only in one header for bound hosts; and changes come back as a patch applied file by file after the signer decides each file. Box fails closed, and each job ends as a .tkb bundle anyone can verify offline. All 13 end-to-end checks pass on a real KVM runner (the test that runs them takes 28.5 s, including image pull and boot); 15 fuzz targets on the trust boundary produced nine findings, all fixed; and the demo run's 102 signed records verify. Box does not yet run Firecracker's jailer, needs a co-located signer, and records processes inside the VM only where they cross a door.
We study the problem of finding the minimum $\mathbb{L}\_2$-norm element in a transportation polytope. It corresponds to minimizing the index of coincidence or maximizing R{é}nyi entropy over joint probability distributions with prescribed marginals. The solution is straightforward under a feasibility condition on the marginals. We make three contributions. First, we decompose the basic solution into a mixture of three interpretable couplings and derive a closed-form sampling procedure, revealing its role as the $\mathbb{L}\_2$-projection of the uniform distribution onto the transportation polytope. Second, we show that the feasibility condition becomes exponentially rare as the number of modalities grows, which highlights the practical necessity of a more general construction. Third, we provide a KKT-based iterative construction of the optimal coupling for arbitrary marginals with explicit Lagrange multipliers.
Jungwoo Oh, Hyunseung Chung, Junhee Lee, Min-Gyu Kim, Hangyul Yoon, Ki Seong Lee, Youngchae Lee, Muhan Yeo 等 9 位作者
原文摘要
While Multimodal Large Language Models (MLLMs) show promising performance in automated electrocardiogram interpretation, it remains unclear whether they perform actual step-by-step reasoning or just rely on superficial visual cues. To investigate this, we introduce ECG-Reasoning-Benchmark, a novel multi-turn evaluation framework comprising over 6,400 samples to systematically assess step- by-step reasoning across 17 core ECG diagnoses. Our comprehensive evaluation of state-of-the-art models reveals a systematic shortfall in executing multi-step logical deduction. Although models possess the medical knowledge to retrieve clinical criteria for a diagnosis, they exhibit consistently low success rates in maintaining a complete reasoning chain, primarily failing to ground the corresponding ECG findings to the actual visual evidence in the ECG signal. These results demonstrate that current MLLMs bypass actual visual interpretation, highlighting a limitation in their visual grounding capabilities and motivating robust, reasoning-centric medical AI. The code and data are available at https://github.com/Jwoo5/ecg-reasoning-benchmark.
Decision tree ensembles are widely used for tabular prediction tasks in judiciary, healthcare, banking, etc. We show that the piecewise-constant structure of these models can create narrow local oscillations, leading the ensemble score to sharply drop and then rise (a downward spike), or sharply rise and then drop (an upward spike), over a small change in a single feature. We call such patterns glitches. A glitch is not, by definition, an error. Rather, it is a formally checkable audit target requiring inspection. Formally, glitches are witnessed by ordered input triples in which a small monotonic change in a single input feature causes the output to first change in one direction and then revert in the opposite direction. Through a substantial experimental study, we show that glitches have widespread existence in even the state-of-the-art models performing various critical tasks. Our findings also indicate that glitches enrich and extend existing notions like robustness and monotonicity, providing us a more refined tool to isolate anomalies for inspection. On the theoretical side, we prove that deciding whether a decision tree ensemble contains a glitch is NP-complete, even for bounded-depth trees. We then present a mixed-integer linear programming encoding for the global and local glitch verification problems, and implement our algorithm in a verifier for gradient-boosted decision trees. Experiments show our verifier can scale to realistic model sizes from the literature.