让下一篇重要论文，不再错过。

Psychological harms from interactions with large language models cannot be monitored directly because they may emerge over repeated interactions and depend strongly on subjective individual user context. We propose measuring risk, not harm: detecting observable AI behaviors that constitute risk factors for psychological harm at the AI response level and aggregating these signals to support longitudinal monitoring. We operationalize eight such behaviors and construct a synthetic multi-label corpus of 47,062 AI responses and 373,873 response-factor labels, with targeted human annotations for validation and evaluation. We then compare trained embedding-based classifiers, direct LLM judging, and a classifier--judge cascade. All detectors are model-agnostic and operate only on LLM outputs, requiring no access to internal model states. We apply these detectors to several out-of-distribution human-AI conversation data sets from prior research. On held-out human-labeled data, the best cascade achieves macro precision of 0.86, recall of 0.76, and F1 of 0.79. On WildChat, it reduces LLM-judge calls by 97.1\% relative to direct judging. On DelusionEval, risk-factor detections cluster within conversations. Together, our results show that response-level behavioral risk-factor detection can provide a scalable measurement layer for longitudinal psychological-safety monitoring without conflating observable AI behavior with latent user harm.

A graph class $\mathcal{C}$ is monadically dependent if one cannot interpret all graphs in vertex-colored graphs from $\mathcal{C}$ using a fixed formula of first-order logic. We prove that every monadically dependent graph class has almost bounded merge-width, thereby confirming a conjecture of the first and the last author [STOC '25]. Our proof is algorithmic and the witnessing decompositions in the form of merge sequences can be computed in nearly-cubic time. This is a step towards the proof of fixed-parameter tractability of model checking first-order logic on monadically dependent graph classes, which is achieved in a companion paper.

World models learn to predict the temporal evolution of visual observations given a control signal, potentially enabling agents to reason about environments through forward simulation. Because of the focus on forward simulation, current world models generate predictions based on factual observations. For many emerging applications, such as comprehensive evaluations of physical AI behavior under varying conditions, the ability of world models to answer counterfactual queries -- such as "what would happen if this object was removed?" -- is of increasing importance. We formalize counterfactual world models that additionally take interventions as explicit inputs, predicting temporal sequences under hypothetical modifications to observed scene properties. Traditional world models operate directly on entangled pixel-space representations where object properties and relationships cannot be selectively modified. This modeling choice prevents targeted interventions on specific scene properties. We introduce CWMDT, a framework to overcome those limitations, turning standard video diffusion models into effective counterfactual world models. First, CWMDT constructs digital twins of observed scenes to explicitly encode objects and their relationships, represented as structured text. Second, CWMDT applies large language models to reason over these representations and predict how a counterfactual intervention propagates through time to alter the observed scene. Third, CWMDT conditions a video diffusion model with the modified representation to generate counterfactual visual sequences. Evaluations on two benchmarks show that the CWMDT approach achieves state-of-the-art performance, suggesting that alternative representations of videos, such as the digital twins considered here, offer powerful control signals for video forward simulation-based world models.

Efficient benchmarking estimates a system's full benchmark performance from a small number of task executions. Standard approaches rely solely on the binary correctness of executions from a collection of already-evaluated reference systems. An agentic system, however, produces a rich trace for each execution that can reveal substantially more about its capabilities. Leveraging this information requires useful trace representations. We propose representing a trace by extracting a structured summary of it with an LLM and embedding the summary with a standard text encoder. These representations preserve correctness- and task-related signal while minimizing the presence of system identity-, model family- and harness-related signals that dominate raw trace embeddings. Given this representation for each task execution from all reference systems, we construct low-dimensional representations of the systems and predict a new system's score from its position among the reference systems. On SWE-bench Verified and Terminal-Bench 2.0, augmenting an item-response model with this geometry-based prediction reduces mean absolute error from 0.13 to 0.08 (Wilcoxon $p < 10^{-4}$) and from 0.14 to 0.12 (Wilcoxon $p < 10^{-3}$), respectively, when provided a single random executed task. Further, given access to the full set of reference systems, our approach dominates baselines for any given number of executed tasks on both benchmarks.

Hamilton-Jacobi (HJ) Reachability identifies the unsafe states of a dynamical system from which it can be forced into failure under worst-case disturbance. However, classical grid-based solvers scale exponentially with state dimension. DeepReach improves scalability by replacing the grid with a neural value function trained on the HJ-Isaacs (HJI) variational inequality (VI), but penalizes both sides of the residual equally. This allows the learned value to lie above the true one and unsafe states to be misclassified as safe. We propose Conservative DeepReach (ConDR) which aims to learn a subsolution of the HJI VI such that the learned unsafe set contains the true one. ConDR enforces the obstacle branch and terminal condition by construction, pushes the HJI residual to a safety margin $ε$ below zero with a one-sided loss, and prevents the degenerate solution of classifying everything as unsafe. We prove that the exact solution of the margin-shifted VI that ConDR trains toward is a subsolution of the original VI so its unsafe set contains the true one and its value lies within $ετ$ of the true value function. We compare against DeepReach and MADR in the established systems of Air3D, Air6D, a 9D multi-vehicle system, and a 13D quadrotor. ConDR reports the lowest rate of unsafe state misclassified and the lowest closed-loop collision rates with at most a small increase in intervention. On Air3D and Air6D, ConDR tolerates model mismatch that scales with $ε$. At matched false-negative rates, ConDR's unsafe sets are smaller than the baselines showing the safety is not obtained by trivial over-approximation.

Message-passing graph neural networks (GNNs) are limited in expressive power by the one-dimensional Weisfeiler-Leman test. Persistent homology (PH) provides topological descriptors that can complement GNN representations, but existing approaches typically compute persistence separately from the representation learning process. In this paper, we introduce Message-Passing Graph Persistence (MGP), a family of topological descriptors that incorporates message passing into graph filtrations. At each filtration step, node embeddings are updated using the current subgraph and associated with persistence events. These embeddings also determine the representatives of connected components during persistence computation. We prove that MGP is invariant under graph isomorphism and that it is strictly more expressive than the PH-based descriptors it extends. We evaluate MGP on an expressivity benchmark and on several graph classification and regression datasets. Experiments show that MGP outperforms other PH-based descriptors on most of these datasets.

Black-box variational inference (BBVI) is a methodology for posterior approximation that relies on stochastic optimization. In practice, the stochastic optimizers underpinning BBVI generally require extensive problem-specific tuning, which undermines its promise as a truly "black box" inference algorithm. However, over the past decade, many new adaptive stochastic optimization algorithms have been developed that reduce or remove entirely the need for tuning. In this work, we investigate this new collection of adaptive methods in the context of BBVI, with the goal of establishing the current state of the art in tuning-free optimization-based inference. In particular, we present a large-scale empirical evaluation of 56 stochastic gradient-based optimization algorithms applied to 1092 Bayesian inference optimization problems, involving over 550,000 individual optimization runs and 15 core-years of compute. The optimization algorithms we evaluate are chosen to represent a wide spectrum of recent approaches and the benchmark problems are chosen to span a range of difficulty, with posterior target dimension 1-10^4, condition number 1-10^8, and a range of variational families. Our results show that no single method dominates, but running a selection of 5 algorithms suffices to reliably get close to the best-possible observed performance. We thus provide a strong baseline for applications where expert tuning is not possible and for comparison when developing new stochastic optimization algorithms.

Ensuring neural networks adhere to domain-specific constraints is crucial for addressing safety and trustworthiness while also enhancing inference accuracy. Despite the nonlinear nature of most real-world tasks, the majority of existing methods are limited to affine (equality) or convex (inequality) constraints. We introduce ENFORCE, a neural network architecture that uses an adaptive projection module (AdaNP) to enforce nonlinear equality and inequality constraints in the predictions up to a specified tolerance $\varepsilon$, and exactly in the affine-in-$y$ case. For affine constraint sets, we prove that the associated projection mapping is non-expansive (1-Lipschitz), ensuring stable gradient propagation. For nonlinear constraints, we establish local convergence analysis under standard regularity conditions. We evaluate ENFORCE on multiple tasks, including function fitting, real-world engineering case studies, and learning optimization problems. For the latter, we introduce a class of scalable optimization problems as a benchmark for nonlinear constrained learning. In the benchmarks, the predictions of our architecture satisfy nonlinear equality and inequality constraints up to a prescribed tolerance $\varepsilon$, while maintaining scalability with tractable computational complexity at training and inference time.

Few-step distillation offers efficient generation with diffusion and flow models, but preserving quality with only a few sampling steps remains challenging. Distribution-matching methods via score distillation provide a strong foundation, yet they rely on accurately tracking the student's changing distribution. At intermediate generation steps in few-step sampling, student-score estimation can be difficult, and commonly used estimators may describe a different distribution from the one used to update the student. Flow matching allows a more flexible approach to supervision through velocity fields constructed from the student's sampling process. Building on this perspective, we introduce Velocity Adversarial Distillation (VAD), which uses a learned critic to compare student and teacher fields and turn their disagreement into feedback for learning. VAD unifies score distillation, Fisher-based matching, and moment matching, also explaining existing heuristic MMD updates and delivering new training objectives. We further develop VA-TDM, which follows trajectory distribution matching: providing supervision for partial denoising along the student trajectory. Beyond this, VA-TDM combines transition-aware online velocity learning and dual-form VAD supervision across intermediate student transitions. Extensive results show that VA-TDM is powerful for learning few-step generators, achieving state-of-the-art four-step performance in both text-to-image and text-to-video generation across various backbones and scaling effectively to 33B MiniMax-H3. In particular, our VA-TDM-H3 notably outperforms the few-step Fast-H3 and is closer to the original many-step H3 according to human preference.

What should data teach a language model at a particular point in training? A circuit view reveals three distinct bottlenecks: forming a computation, making its required content available, and selecting among available routes. A shared diagnosis-to-data principle connects them: localize the missing operation, preserve its causal relation, vary shortcut-bearing context, and re-audit the residual. Formation-sensitive selection and prerequisite ordering accelerate a binding-matching-transport path; a brief early prefix from the same training multiset retains a validation advantage through 100B tokens. Availability counterfactuals then distinguish writing content from invoking available memory, while paired supervision and context-opportunity ranking improve matched route decisions and long-context answer likelihood. A continuous 350M-model experiment connects the three interventions on the same facts: early circuit training improves subsequent learning, and the complete sequence outperforms stage-replacement controls on facts withheld from Use teaching. Independent query surfaces and opposed-source decisions expose conditional arbitration as the remaining frontier. Together, these results show why a change in the limiting operation calls for a change in supervision, not merely a new ranking of difficult examples.

Sample-efficient fluid intelligence requires agents that rapidly form working models of unfamiliar interactive environments. ARC-AGI-3 tests this across uninstructed visual games scored by Relative Human Action Efficiency, which penalizes exploratory waste quadratically. When frontier models read raw integer grids, serialization destroys spatial topology, and they spend hundreds of actions rediscovering basic boundaries and symmetries. We introduce SARBLOH, an open-weight agent built on the premise that spatial perception, not planning capacity, governs interactive reasoning on ARC-AGI-3. A zero-parameter O(HW) perception engine turns each frame, in 35 ms, into an object inventory with bipartite temporal tracking, a geometric briefing that separates proven invariants from affordance hypotheses, and an annotated canvas. Paired with a code-as-action runtime, the agent writes algorithmic hypotheses instead of low-level moves. Across eight presentations, grounded perception raises the rubric score from 3/28 to 25/28 points while cutting token use by 89.4%. On 23 held-out games, geometric invariants generalize with 100% accuracy, while single-frame affordance guesses fall to 6-50%, so active verification probes remain necessary. On the public set, Run A (briefing off) solved 143 of 183 levels in 8.2 hours at 59.4% single-pass efficiency; Run B (briefing on) reached 63.5%. Within the nine-hour budget on one Kaggle RTX PRO 6000 GPU, SARBLOH scored 16.53 on the hidden Kaggle evaluation, 13.7x the first Milestone winner.

High-quality teleoperation datasets are costly to collect, particularly for hard tasks. We observe that many tasks exhibit directional asymmetry: completing the forward hard task is difficult, whereas reversing it by relaxing or disrupting the environment is comparatively easy. This suggests that reversed easy-task trajectories can serve as a scalable supervision signal for the hard task, reducing the cost of manual demonstration collection. However, reversed data can be noisy, and directly training on it may yield suboptimal policies. To enable largely automated acquisition and effective use of reversed data, we propose a teleoperation-cost effective framework for hard policy learning via temporal reversal of easy tasks, consisting of three key components: a closed-loop data collection pipeline that alternates between hard-task and easy-task policies to autonomously reset the environment and generate diverse trajectories; a hierarchical data refinement pipeline that temporally inverts easy-task rollouts and filters low-quality motion using kinematic priors and a critic-guided advantage filter; and an iterative policy learning method that trains the hard-task policy using both initial reversed easy-task demonstrations and the filtered reversed data in a continuous online learning loop. By combining automated collection, hierarchical refinement, and iterative learning, our method enables scalable, reliable training of complex, high-precision manipulation tasks. Across two simulated benchmarks and real-robot experiments, we demonstrate that our method improves hard-task success rates with higher data efficiency and more stable training compared to reversal-based and reinforcement-learning baselines, without requiring extensive hard-task teleoperation.

In the cut-query model, we have access to a (directed) graph via an oracle and we can query the size of the (directed) cut of a given subset of the vertices. One of the most elementary tasks in this model is to decide if there is a path between two fixed vertices $s$ and $t$. While many results are known for undirected graphs, much less is understood for directed graphs in the cut-query model. Even for the basic task of $s$-$t$ reachability, the best known randomized algorithm is to reconstruct the entire graph with a technique by Grebinski and Kucherov using $O(n^2 / \log n)$ queries. On directed acyclic graphs (DAGs), we obtain a deterministic single-source reachability algorithm using $O(n \sqrt{n \log n})$ queries. The result is based on a topological sort algorithm and can also be adapted to compute single-source shortest paths in DAGs. On general directed graphs, we obtain deterministic algorithms for single-source reachability, strongly connected components and minimum $s$-$t$ cut using $O(n^{9/5}\log n)$ queries. The algorithm is simple and relies on two natural greedy subroutines and a modified version of Dinic's algorithm.

Objective: To develop the Cardiologist Annotation Tool (CAT) for frame-level cardiac view and image-quality annotation in fetal echocardiography and compare it with spreadsheet-based annotation. Methods: CAT provides cine playback, frame navigation, and keyboard-based labeling without manual frame-range entry. Experts annotated 25 studies using both methods, and spreadsheet annotations were audited for predefined data-entry errors. In a randomized crossover study, six medical students annotated two normal studies, one with each method, against an expert reference. Results: Among 1,746 spreadsheet-annotated files, 6.2% contained major or critical data-entry errors, more frequently in congenital heart disease than normal studies (10.9% vs 3.4%). CAT yielded more annotated view segments per clip (3.6 vs 1.5). Rates of labeled and correctly labeled frames per minute were 1.77 and 1.68 times those with spreadsheets, respectively (both P = 0.031), with gains in all six students. No detectable difference in view-label accuracy was observed. Conclusion: CAT eliminates manual frame-range entry and enables finer-grained, structured annotation of fetal echocardiography, providing a practical approach to preparing datasets for AI research.

Organisations deploying AI face a scoping problem: which tasks can be automated, which should remain with humans, and which are best shared between the two. Aggregate benchmark scores provide little insight into where systems will succeed or fail in practice, while human judgements of model capabilities quickly become outdated. We introduce a pipeline that profiles agents and tasks using a shared set of core cognitive capabilities. Cognitive capability profiling infers an agent's capabilities from performance on a benchmark battery annotated for the cognitive demands of each item. Task requirements weighting elicits from domain experts the relative importance of these same capabilities for their work. As both use a common set of cognitive dimensions, they can be updated independently as models and roles change, and combined to estimate AI suitability at the level of a domain, organisation, role, or individual duty. We validate capability recovery on synthetic agents, profile six AI systems, and elicit task requirements from 410 employees across six occupational domains. AI systems differed more across cognitive dimensions than across model families, while workplace activities converged on a shared cognitive core. The resulting scores provide a comparative scoping tool for identifying promising candidates for piloting and areas where current systems are unlikely to be well suited. We discuss extending the framework to profile human workers alongside AI systems, moving from AI suitability towards human-machine task allocation.

Most vision-language models (VLMs) treat the vision encoder as a static feature extractor: it encodes an image once, independently of the question, and leaves subsequent reasoning to the language model. Yet different questions may require different visual operations and processing depths. We present AdaLoop, a framework for question-conditioned visual recurrence in VLMs. AdaLoop adapts visual computation to the question and reuses the encoder's transformations across multiple iterations, with the pretrained vision weights frozen. We also introduce VisRO, a procedural dataset and generation pipeline for training and evaluating vision-centric reasoning. On six VisRO tasks, AdaLoop outperforms standard fine-tuning and question-conditioned single-pass baselines with matched training data and optimizer updates. These gains hold across three VLM families on both in-distribution and out-of-distribution examples. Our findings suggest that pretrained vision encoders can contribute to reasoning through task-directed, iterative visual computation.