Papers

Do AI assistants help believers reason about moral dilemmas consistently with their faith? We present FaithfulBench, the first benchmark to score AI counsel across traditions by how well it adheres to the user's professed faith. Scenarios are drawn from each tradition's most respected texts, with the faithful answer known and applied by the judges as the standard. We test five frontier models under three conditions: the AI does not know the user's tradition; it receives a one-line prompt identifying the user as a practicing adherent; or it receives a companion-counselor guide rooted in the tradition's sources. Two judges score the initial response and whether the model caves or holds when pressured toward the answer the user wants. When the tradition is unstated, models counsel from a secular therapeutic default and every model fails some believers. Naming the faith wins a faithful first answer but not steadfastness; the guide improves both.

Symmetric rendezvous on the line is a search problem in which two agents, initially placed at distance $2d$, must follow the same randomized strategy to meet as quickly as possible. In the standard model, agents may condition their actions on the entire execution history, and both the known- and unknown-distance variants admit expected rendezvous time $\Theta(d)$. We study the role of memory by introducing oblivious self-distance strategies, in which an agent's decision depends only on her position relative to her own starting location. For an initial separation of $2d$, let $R_d$ denote the optimal oblivious expected rendezvous time in the known-distance setting. We develop two finite-state frameworks based on absorbing Markov chains. Truncated chains give computable upper bounds through finite-support strategies, while weak-peek chains give lower bounds through a revealed-information relaxation. Together, they provide a mechanism for certifying optimality. Using that mechanism, we determine $R_1$ exactly and prove that it is attained by a finite-support strategy. For $d=2,\ldots,6$, numerical optimization gives the same truncation structure and objective values, yielding rigorous upper bounds below $7.83d^2$. We do not prove that the computed weak-peek minimizers are global, but the stability of the computations leads us to conjecture that they are, in which case the corresponding truncated strategies are optimal. We also prove that $R_d=\Theta(d^2)$. In the unknown-distance setting, we construct a universal strategy, independent of $d$, with expected rendezvous time $O(d^{2+\eta})$ for every fixed $\eta>0$. Thus, under the memory restriction, the known-distance rendezvous time becomes quadratic, while near-quadratic performance remains possible even without knowing $d$. The asymptotic analysis uses birth-death Markov chains and their electrical-network interpretation.

Let $R^\ast(n)$ be the minimum radius of a central circle externally tangent to nonoverlapping circles of radii $1,\ldots,n$. We prove that $R^\ast(n)=C_\ast n^2+o(n^2)$ and give finite linear programs that enclose $C_\ast$ with prescribed error. An exact angular comparison reduces the problem to marked points on a line with separation at least the geometric mean of each pair of marks. Concatenation with bounded joining cost and reassignment to the prescribed normalized radii prove existence of the limit. Balanced finite-word programs give a bracket of width $(1/k+1/r)/\pi$, before directed arithmetic error, using $k$ types and words of length $r$. We also obtain explicit geometric bounds. For the upper bound, we recover finite and countable reflected-block constructions by actual permutations, retaining all pair constraints. A coupled terminal-subset argument gives an explicit lower bound strictly above $1/8$. Together these give $C_{\rm term}+\eta_{\rm width}\le C_\ast\le U_4$. The lower endpoint comes from the terminal-subset argument, and the upper endpoint from the four-block construction; neither is asserted sharp. The finite-word scheme establishes computability; efficient high-precision evaluation and the structure of global optimizers remain open.

Many labor-intensive tasks in fruit production such as pruning require physically interacting with the plant (e.g., pushing, pulling, bending limbs, etc.). Due to increasing labor shortages, there is widespread interest in the adoption of robotics in this area. When the robot must physically interact with the system, a deformable model of the plant is beneficial. Coupled, spring-loaded beams have been used in prior work to build deformable plant models in simulation for learning, planning, and control, but this approach has rarely been validated against the mechanical properties of real-world plants. In this paper, we use this rigid-body model, grounded in real material properties and beam bending theory, to simulate the deformation of blueberry canes under loading. To validate the model, we simulate the canes in MuJoCo and compare their behavior with real-world data collected from probing canes at a commercial farm with a custom testbed. We perform sensitivity analysis on several key modeling variables and show that this approach is highly sensitive to the plant diameter calculations and a priori flexural modulus estimation, which is dependent on season and blueberry variety. We also share our dataset of live blueberry cane deformation, including RGB-D images and measured forces and displacements.

Large Language Models (LLMs) are powerful zero-shot learners but remain prone to misalignment with human preferences, often producing biased, toxic, or otherwise harmful outputs. Existing alignment methods, while effective, are costly and tightly coupled to the model, limiting flexibility and scalability. We propose a modular correction framework that augments pretrained LLMs with Activated LoRA (aLoRA) adapters and a context-aware routing mechanism to eliminate harms from misaligned model responses. Our approach enables expert adapters to activate mid-sequence without invalidating the KV cache, allowing low-latency, targeted correction during generation. Each expert is trained to detect and mitigate specific harms, such as bias or toxicity. A learned router dynamically selects appropriate experts based on the models intermediate outputs. We demonstrate that our system improves alignment on standard safety benchmarks while preserving task performance, offering a lightweight and efficient path toward safer and more controllable LLM deployments.

This paper studies joint sampling, route activation, and preemption control over $N$ heterogeneous transmission routes under a long-run objective combining Age of Information (AoI) with sampling and transmission costs. We formulate the problem as an average-cost impulse-control model, derive a vector-form integral average-cost optimality equation over a hybrid state space, and establish structural reductions on the state. We consider a two-route case and compute optimized policies using discretized policy iteration. Surprisingly, the results show that preemption can reverse routing preferences: the optimized policy may exclusively use a route with both a larger mean delay and a larger delay variance while avoiding the seemingly superior route. Relative to baselines that do not allow preemption, the proposed policy reduces average AoI by up to $96\%$.

OpenAI released ChatGPT Images 2.5 on 8 September 2026, advertising more precise local edits, better consistency across edits, more faithful reference products and sharper detail. We evaluate these claims on four forgery tasks with answers fixed in advance: receipt-field alteration, repeated editing, product placement and small-print rendering. GPT-Image-2 provides same-week baselines at a cheaper and a more expensive tier. A limited improvement appears in receipt editing. After alignment, OCR detects changes to surrounding text in 31.7% of Flare outputs, against 44.2% for the cheaper baseline. This gain is concentrated on CORD receipts and sensitive to shifts of a pixel or less; the forged value itself is no more often correct. Repeated editing and fine print show no measurable gain. Product codes become more legible mainly because Images 2.5 draws the product larger. Defence outcomes change little: localisation remains weak for both generations. A detector that flags 68.6% of controlled benchmark images flags only 35.9% of images posted online. Advertised improvements therefore transfer unevenly to the tested forgery capabilities, while substantial detection limitations remain.

For our submission to the WMT26 Creole Language Translation Shared Task, we focus on machine translation (MT) models for Pacific creoles: Tok Pisin, Bislama, and Solomon Pijin, with particular attention to broad domain performance. After pre-training on a large collection of domain-imbalanced data, we continue fine-tuning on a diverse mix of domain-balanced data. We rely on a number of data collection and preparation techniques, including LLM-assisted respelling and alignment, back-translation, and distillation from Gemini for domains originally not present in training data. Evaluated on Bouquet and a novel test set made of spoken language transcripts, our models beat open model baselines by 3+ chrF++ points in all directions with human-original references. Looking ahead, we plan to develop human-translated test sets for Solomon Pijin and Bislama, and to distil our best models into much smaller ones that retain broad domain coverage.

Generative AI is currently fragmented, with separate subscriptions required for text generation, image synthesis, and code assistance, leading to high costs and workflow disruptions. AI Express is a SaaS platform that consolidates these capabilities into a unified pay-as-you-go service. The platform provides a range of AI tools, including a Core Code Reviewer that evaluates GitHub repositories, a Text-to-Website generator that converts text into fully functional websites, script generation, grammar checking, data analysis, and watermark removal, all accessible through a single interface without requiring technical expertise.

AI sycophancy, the tendency of AI systems to excessively agree with, flatter, or validate users, is an emerging concern in human-AI interaction. It is especially consequential in problem-sharing contexts, where users may seek both information and emotional validation. This paper conceptualizes AI sycophancy as excessive response validation and introduces a five-level AI Response Validation Framework (ARVF), measured using a six-item Perceived AI Sycophancy Scale (PASS). Across three phases, the study validated the framework with human participants through an online questionnaire, tested LLM-as-evaluators in answering PASS, and examined how ten contemporary LLMs generated and evaluated responses to real-world work and personal conflict scenarios. Results supported the intended progression of perceived sycophancy in ARVF and the reliability of PASS. Trust and perceived competence followed an inverted U-shaped pattern. LLM ratings reproduced the five-level structure but showed calibration differences from human ratings. Based on the LLM ratings, the ten contemporary LLMs were grouped according to their behavioral pattern. Text-based analysis provided further support on the linguistical structure for sycophantic responses generated by LLMs. Findings highlight AI sycophancy as a graded, context-sensitive interactional phenomenon.

Symmetry reduces the capacity of a quantum learning model, but the imposed group must match both the measured information and the label transformation. We establish a finite-measurement theory for inferring this group from candidate transformations. The central structural result identifies observable-invisible transformations with the stabilizer of a projected state whenever the probe span is invariant. It turns recovered generators into a valid subgroup and identifies the continuous invisible space with its Lie algebra. For finite dictionaries, an unbiased shadow statistic distinguishes zero from positive squared expectation discrepancies with an inverse-gap measurement rate, improving the inverse-square-gap rate of uniform discrepancy estimation. A commuting qubit lower bound proves the gap dependence optimal at fixed snapshot scale, and simultaneous intervals support data-dependent tolerances. Task validation then tests either the joint distribution through a characteristic kernel or its encoded mean through a classical–quantum discrepancy. An exact group-average identity relates the latter to joint-state asymmetry and specifies its conversion to binary task breaking mass. Projection bias quantifies the cost of excessive symmetry, while an $\ell_1$ readout bound quantifies the capacity gained by relaxing it. At an invariant pure-state backbone, retained and nontrivial breaking sectors are Fisher-orthogonal. Ising-chain calculations connect finite-shot recovery, label-dependent symmetry, and physical sector drift. These results determine which symmetry the measurements support and provide the statistical and geometric basis for a subsequent release decision.

Scalar rounding decisions interact through matrix multiplication. We study deterministic product-aware rounding after scales, clipping bounds, and grids are fixed, with each active scalar choosing between adjacent levels. For dynamic activation rounding, null-space reduction preserves the relaxed product while leaving at most $r$ fractional decisions, where $r$ is the rank of the active gap-weighted weight block. Conditional-expectation completion gives a deterministic polynomial-time algorithm with squared product error at most $ \mathrm{OPT}_{\mathrm{dyn}}+r\nu_{\max}^2/4$, where $\mathrm{OPT}_{\mathrm{dyn}}$ is the best admissible error and $\nu_{\max}$ is the largest row norm of that block. For reusable static weights, the exact expected product-loss metric is the uncentered input second moment with fixed output bias; free bias recalibration yields the centered covariance. Exact optimization is NP-hard even at rank one. In balanced blocks with $K=1024$ and $r=16$, conditional- expectation completion attains a dither-normalized median error of $0.010$, compared with $0.899$ for round-to-nearest. Clipping-aware initialization reduces median normalized error by a factor of $43.4$ at ten-percent clipping. Held-out Digits experiments show that retaining the input mean or correcting the output bias improves median product error over round-to-nearest in all four tested bit-width and calibration-size settings.

Continual Learning Bench scores learning as short-horizon gain versus a reset baseline and finds naive full-context ICL strongest among the memories it tested. We treat ICL as one learning system and score it on a longer shared-world schedule. Steady-state learning is the gap versus baseline on a pre-set late window (last 40 of 174 BIRD-SQL formula-1 questions). We split the score into exploration efficiency (SQL probes), task reward (hits), and delivery cost (API dollars and context size). On gpt-5.6-luna, late probes fall from 4.6-5.6 to 0.95 while hits rise only modestly and ICL context grows to about 95k tokens with cost roughly doubling. Short-horizon gain understates the late probe saving and misses the cost inversion, so we find that unbounded ICL is a poor candidate for the learning mechanism.

Diagnosing rolling-element bearing faults from vibration is a canonical physical-sensing task and a widely used benchmark for domain adaptation under operating-condition shift. Accuracies above 99 percent are commonly reported, but under evaluation splits that place the same physical bearing in both training and test. We revisit the task under a held-out-bearing protocol, assigning every bearing unit entirely to either the training or the test set, and find that source-only transfer is far weaker than such numbers suggest: on a change of shaft speed it reaches only $0.36$, against a target-supervised ceiling of 0.97. We then study what governs transfer. Treating computed order tracking, a shaft-angle resampling that places fault frequencies at fixed shaft orders independent of running speed, as a controlled change of representation, we find that a Fourier Neural Operator raises source-only transfer from $0.36$ to $0.61$ on the speed shift, where the fault peaks move, while a convolutional network of matched feature dimension stays near chance in both representations. The representation also decides whether unsupervised alignment can work: with the same normalized RBF-MMD loss and no target labels, the operator reaches 0.71 in the frequency domain but 0.95 in the order domain, within 0.02 of the target-supervised ceiling and above $0.86$ on every held-out bearing fold. Once the representation is right, a small label budget adds little. These results indicate that, for this task, the input representation rather than the alignment method decides whether adaptation helps. A second dataset, whose held-out units are fault diameters rather than bearings, shows that the same protocol exposes failures that even a target-supervised model cannot avoid.

Modern design optimization frameworks aim first and foremost for models with the most accurate predictions without balancing computational overhead. It remains a reason why scaled architecture and multidisciplinary design optimization problems are difficult to address, even with sample-efficient Bayesian optimizers. In this paper, a metric quantifying the computational energy footprint is introduced within a Bayesian optimization framework to guide the parameter setting of a model towards configurations that balance both performance and frugality. The computer experiments highlighted existing tradeoffs between optimum convergence and the underlying energy footprint, and sometimes resulted in both a better-found optimum and lower energy consumption.

The purpose of this article is to highlight the central role of autonomous systems as the ultimate stage in the development of AI, to explain the underlying technical challenges that require a combination of connectionist AI and symbolic AI, and to integrate AI and systems engineering. We present a comprehensive framework for the design and evaluation of autonomous systems, based on a generic agent architecture that characterizes their behavior as the composition of cognitive functions organized around a long-term memory containing the agent's evolving knowledge. We address the challenges posed by the implementation of the fundamental features of the agent architecture, in particular the link between sensory data and structured data stored in memory, decision-making related to the achievement of the agent's goals and their planning, as well as the coordination of agents to combine individual and collective intelligence. We explain that agent trustworthiness, unlike that of traditional systems, is not limited to behavioral properties. It includes an essential dimension related to cognitive properties, the validity of which depends on how the agent uses its knowledge in decision-making. We present avenues for the development of methods for evaluating agent trustworthiness. We conclude with a critical assessment of the substantial gap between the aspirational vision of autonomous multi-agent systems and the current state of the art.

In an approval-based committee election, a committee of size $k$ is selected from a set of candidates to represent voters of total weight $n$, each of whom has positive weight and approves a subset of the candidates. A size-$k$ committee is $\lambda$-stable if, for every nonempty subset $T$ of candidates, the total weight of voters who strictly prefer $T$ is less than $\lambda$ times the proportional share $n|T|/k$. Whether there exists a committee that is exactly stable, corresponding to $\lambda=1$, remains a major open problem in approval-based committee voting. Thus, a natural objective is to identify small values of $\lambda>1$ for which $\lambda$-stability can always be guaranteed. We prove that weighted sequential Phragmén, a classical and natural rule, always returns a $2$-stable committee of size $k$. Since the $\lambda$-core is the set of all $\lambda$-stable committees, our result implies that the $2$-core is always nonempty. This improves upon the previously best-known guarantee, due to Gao, Sun, and Vondrák~[EC~2026], that a $3.651$-stable committee always exists.

We study exact representability by bias-free shallow polynomial neural networks using algebraic geometry. Over $\mathbb{C}$, a width-$r$ network with activation $z\mapsto z^d$ computes a sum of $r$ $d$-th powers of linear forms, whose Zariski closure is a Veronese secant variety. Ideal elimination therefore yields polynomial certificates of nonrepresentability. We implement this construction as a generic architecture-to-certificate pipeline. For quadratics, we recover the exact symmetric determinantal description and explain its dimension through orthogonal symmetry. In higher degree, the implementation recovers classical catalecticant and secant equations and maps the practical reach of direct elimination across a finite architecture sweep. We also derive the exact population loss floor for a rank-two quadratic network on the sphere, illustrating how an algebraic obstruction induces irreducible approximation error.

The Dynamic Window Approach (DWA) is widely used for local navigation, but its performance depends strongly on parameters that are typically fixed before navigation. In particular, the appropriate prediction horizon can vary with local free space: longer horizons support efficient motion in open areas, whereas shorter horizons help preserve feasible motions in narrow or cluttered regions. This paper proposes D3DWA, an adaptive DWA framework based on a Dueling Double Deep Q-Network (D3QN), which jointly selects the DWA evaluation weights and prediction horizon from a continuous navigation state at every control step while retaining DWA's trajectory generation and collision checking. In eight simulated environments, including unseen layouts, D3DWA reached every goal. Real-robot experiments further showed that D3DWA completed all three tested configurations, including a constrained case in which the weights-only variant timed out. These results demonstrate the benefit of jointly adapting the evaluation weights and prediction horizon. Additional material is available at https://mertcookimg.github.io/d3dwa/

Recent advances in humanoid robotics have produced diverse skills through reinforcement learning, motion imitation, and generative modeling. Yet these capabilities remain siloed because they are built around incompatible representations, interfaces, and controllers. We present CHOREO, a framework for training-free composition of heterogeneous humanoid skills. Our key observation is that, regardless of how a skill is learned, it can ultimately be expressed as an executable motion trajectory. Based on this observation, CHOREO converts each capability into SkillMotion, a unified representation that combines motion states, contacts, semantics, and boundary conditions. Skills are composed through direct continuation, cubic Hermite blending, or validated bridge motions, without retraining source models or updating models at test time. On Unitree G1 in MuJoCo, CHOREO organizes 2,950 admitted SkillMotion assets derived from heterogeneous sources and achieves 95.4% sequence success across 130 multi-action tasks, including 93.8% success on eight-action sequences. These results demonstrate that executable trajectories provide a scalable interface for accumulating and composing pretrained humanoid capabilities.

We formulate Artistic Intelligence as exploration driven world realization, leaving space for creative possibility while preserving the semantic, artistic, and compositional structure that must remain true. Moonworks Lunara, a text-to-image model, implements this framework with a novel Diffusion Mixture Transformer architecture. A new training algorithm iteratively evolves the data distribution through informative sample acquisition and targeted injection of human-created art. We benchmark Lunara against seven image-generation models, including FLUX.2-Klein-4B, Qwen-Image (20B), and GPT-Image-1-Mini. With GPT-5.6 Sol as evaluator, Lunara ranks first in Aesthetic Quality (8.473 vs. 8.457 GPT-Image-1-mini), second in Emotional Resonance, and remains competitive in Content Integrity. A blind human evaluation over the same evaluation set corroborates the automated metrics, ranking Lunara first. It also stays among the strongest models under conventional measures including CLIPScore and LAION Aesthetic Predictor. On GenEval, Lunara achieves competitive performance against a broader set of 16 models, including GPT Image 2 and Seedream 4.0. These results place Lunara at the frontier with Artistic Intelligence while maintaining a sub-10B active-parameter footprint and sub-10-second inference latency. Lunara advances the general visual intelligence frontier by shifting the question from whether models can get images right to how deeply they can interpret meaning and realize it as imaginative, expressive worlds.

As AI systems evolve into personalized digital companions, a central capability is reasoning over a user's long-term personal history: not merely storing past events, but tracking longitudinal experiences and evolving preferences. Progress here is bottlenecked by evaluation, existing long-term memory benchmarks are largely synthetic and text-only, they overlook the visual records that anchor everyday human memory, lack the authentic and causally connected longitudinal data that real personalization demands, and consequently remain confined to shallow factual recall. We introduce ReaLMem (Real-world Long-term Multimodal Memory), the first benchmark built from authentic multi-year personal visual archives, paired with first-person subjective annotations. ReaLMem evaluates models across three cognitive tiers of increasing difficulty: factual recall, persona inference, and predictive personalization. We further propose ChronoProfiler, a temporal-weighting profiling module that computes temporal stability scores for user attributes and applies them as a salience prior, resolving conflicts among temporally inconsistent preferences and helping models compound multiple co-active preferences in complex personalized decisions. Extensive evaluation of frontier multimodal large language models (MLLMs) and memory systems on ReaLMem reveals predictive personalization as a consistent ceiling, exposes clear performance gaps and bottlenecks between MLLMs and memory systems, and shows that high-quality, temporally informed representations substantially improve personalization. Together, ReaLMem and ChronoProfiler provide an authentic testbed and a simple, effective mechanism for long-term personalization, laying a foundation for future research on lifelong AI companions.

Evolutionary Ensemble Search (EES) constructs machine-learning procedures through expert-guided program evolution. A role-specialized council turns task evidence and experimental results into structured search directions. An orchestrator allocates these directions to execution specialists and an evolutionary engine. The engine selects measured parents, diagnoses their errors, and produces descendants through code mutation, structured pipeline edits, and crossover. Each child must execute and acquire its own validation evidence. Population archives retain useful alternatives, while compatible predictions compete in a validation-gated ensemble stage. Search adapts through parent-relative operator credit, session memory, and problem-indexed lessons retrieved across runs. We specify these mechanisms, distinguish their execution profiles, and define the contracts required to compare candidates as their computations change. A public MLE-bench Lite development ledger records medal-threshold artifacts on 19 of 22 tasks (86.36%), with best outcomes of 11 gold, five silver, and three bronze. The procedures span text, images, tables, audio, scientific geometry, and deterministic transformations. The campaign includes grade feedback between runs, external-source routes, and mixed confirmation procedures; its aggregate is an achieved development result, not a blind autonomous-agent success rate. The report contributes a concrete architecture for cumulative executable search and a versioned account of its cross-modal development outcomes.

Visual torque feedback supports underwater bilateral teleoperation, but the benefit of mixed reality (MR) over conventional monitor presentation remains unclear. We present MR-GLi, an MR interface that spatially registers a reaction torque indicator and wrist-camera image to the robot gripper. Twenty participants performed lift and pick-and-place tasks with rigid and compliant objects in a counterbalanced within-subject comparison with a 2D monitor, using identical visual-feedback content and four-channel bilateral control. MR-GLi provided gripper-linked access to visual feedback while maintaining a similar level of torque-regulation performance to the 2D monitor. Subjective evaluation further indicated reduced perceived burden associated with shifting attention between the workspace and visual feedback. These results demonstrate the feasibility of gripper-linked MR overlays for underwater bilateral teleoperation and highlight the importance of considering information access in addition to task performance. Additional material: https://mertcookimg.github.io/mr-gli/

Bilateral control-based imitation learning captures both position and force information, making it well suited to contact-rich manipulation. However, existing approaches provide limited means for an operator to specify how a learned task should be executed at inference time, such as slowly or quickly, gently or firmly. We propose Bi-MoDe, a modifier-conditioned decoding framework that injects a constrained latent into every layer of the Transformer action decoder via adaLN-Zero, allowing behavioral directives to directly influence action-chunk generation. We evaluate the method on a real-world whiteboard wiping task with combinations of temporal and physical modifiers. Bi-MoDe improves physical directive following over the action-chunking baseline while maintaining comparable temporal control. An ablation further shows that decoder conditioning and latent-space composition interact, and that their combination is important for accurate physical directive following. Additional material is available at the https://mertcookimg.github.io/bi-mode/

We developed DAMM-Net++, a 2.5D architecture for thoracic OAR and target volume segmentation that addresses three persistent challenges in radiotherapy auto-contouring: inter-slice surface incoherence, systematic failure on small low-contrast targets, and the absence of per-case reliability signals. The central component is an anatomy-change-aware bidirectional selective state-space memory that models through-plane anatomical change and selectively propagates context along the axial slice sequence. A boundary-aware decoder sharpens near-surface predictions, and an uncertainty head provides calibrated per-voxel confidence for clinical triage. We evaluated 2,146 patients across four centers, an independent external cohort of 112 patients, and a multicenter reader study involving 17 radiation oncologists on 305 cases. The model achieves a mean Dice of 0.955 and HD95 of 3.78 mm, with the largest gains on low-contrast organs-at-risk (OARs) and target volumes where through-plane context is most critical. The uncertainty head is well-calibrated and supports case-level triage. In the reader study, AI assistance reduced contouring time by 75-80% across experience levels and raised junior-reader IoU from 0.861 to 0.925, matching the unedited model. External validation showed a modest internal-to-external drop (< 5%) with calibrated uncertainty transferring without recalibration. The complete deployment pipeline from DICOM ingestion to TPS-compatible RTSTRUCT export has been integrated into the clinical workflow at a partner hospital, where it is used to assist with contouring. These results suggest that anatomically motivated inter-slice memory, paired with uncertainty-guided review, offers a clinically viable path for thoracic auto-contouring.

Intracranial arterial calcifications (IACs) are a common finding on clinical non-contrast enhanced head CT scans and are associated with neurovascular disease. Calcifications can occur in the intimal or medial layer of the arterial wall, subtypes that differ in aetiology and may have distinct clinical relevance. These subtypes can be visually distinguished by radiologists based on the shape of the calcifications. We investigate three automated approaches for subtype classification of IAC from head CT-derived segmentation masks: (1) an automated adaptation of the established radiological visual score, (2) a sphericity-based method, and (3) a method based on shape embeddings extracted by a medical shape foundation model. All approaches use the same lightweight classification pipeline on top of the features they compute and are evaluated using 5-fold cross-validation. The three methods achieved comparable performance, with the embedding-based approach yielding the best overall results with a weighted F1 (mean $\pm$ SD) of up to 71.5 $\pm$ 3.7 for a single artery and 59.8 $\pm$ 1.7 for the joint artery classification. Performance was largely preserved when using automated instead of manual IAC segmentation masks, and we found the difference in weighted F1 not significant. Our results show that fully automated IAC subtype quantification from head CT is feasible and remains robust to the use of manual and automated IAC segmentation masks. Code at https://github.com/bjin96/iac-subtyping.

We introduce StepAudio 3 Music, a large-scale, long-form music generation model that supports explicit musical planning and open-domain text-controlled generation. The StepAudio Music Tokenizer represents audio as a 50-Hz stream from a 65536-entry single codebook, using semantically informed self-supervised and multi-task training to preserve musical structure and reconstruction-relevant information. A flow-matching diffusion Transformer (DiT) predicts continuous StepAudio VAE latents, which our VAE decoder converts into 48-kHz audio. This discrete-continuous design is guided by comparisons of single-codebook VQ, Semantic and Acoustic RVQ, and different DiT configurations. For explicit planning, a Mixture-of-Experts autoregressive model uses ABC notation to produce an intermediate arrangement plan (ABC-CoT) before predicting music tokens, making harmony, rhythm, and melodic structure part of the generation context. A progressive training curriculum and supervised fine-tuning support song and instrumental generation, accompaniment generation from dry vocals, and cover-song synthesis for up to 5 minutes and 30 seconds. With reinforcement learning via direct preference optimization (DPO), the final model achieves the highest AudioBox Content Enjoyment, Content Usefulness, and Production Quality scores and the highest MuQ-MuLan similarity among the evaluated systems, with competitive SongBench results. On the preliminary Artificial Analysis Music Arena Vocals leaderboard, it obtains a Quality Elo of 1105, behind only Suno V5.5 and Mureka and ahead of Suno V5, MiniMax models, and other systems. Audio demonstrations are available at https://stepaudiollm.github.io/step-audio-3-music.

Cervical vertebral maturation (CVM) assessment plays a pivotal role in orthodontic diagnosis and determining the optimal timing of treatment, especially for pediatric patients. In this paper, we propose LM-PCVMNet, a novel deep learning framework for automatic pediatric CVM staging. Specifically, our method integrates vertebral anatomical landmark information, heatmap-guided feature modulation, and metadata-informed similarity modeling into a unified learning framework. We introduce a heatmap-guided feature modulation module that enhances feature extraction by leveraging landmark-centered heatmaps to highlight morphologically relevant vertebral regions. A vertebral landmark-prompting block is designed to incorporate anatomical geometry into the representation learning process. Furthermore, we develop a learnable metadata supervised contrastive loss that adaptively modulates positive-pair similarity based on metadata similarity, enabling the model to learn more biologically consistent and discriminative features. To facilitate further research in pediatric orthodontic treatment, we additionally release PCVM+. It contains 1800 lateral cephalometric radiographs from real-world patients aged 3-15 years, with expert-annotated CVM stages, 13 vertebral anatomical landmarks, and corresponding metadata. We perform comprehensive experiments on two datasets, and the results show that our method achieves state-of-the-art performance, effectively improving landmark localization and classification accuracy over existing models. Code and dataset will be available at github.com/ybupengwang/LM-PCVMNet.

Diffusion-based artifact detectors score whole-slide image patches by reconstruction error under a model fine-tuned on clean tissue. We show that conditioning this fine-tuning on random Gaussian embeddings – resampled at every step from approx. 200 KB of precomputed embedding statistics, with no encoder, no cache, and no change to inference – consistently widens the clean/artifact separation. A four-step ablation chain shows the benefit requires neither content (shuffled real embeddings), provenance (synthetic Gaussians), a tuned intensity (flat across an 8x variance range), nor per-patch identity (fresh per-step noise); a LoRA-dropout control shows the conditioning pathway specifically, not generic weight perturbation, carries the effect. Patch-level gains of +0.25-0.48 Cohen's d replicate across nine trainings; honest leave-one-slide-out evaluation clears a pre-registered bar in 2/2 seeds; and two pre-registered external endpoints on a 281-case set confirm pooled Delta F1 = +0.0073 (95% CI) and +0.0129 (97.5% CI, two-look corrected). We release the full evaluation protocol, including measured seed noise and selection-optimism pricing.