Artificial Intelligence
Causal Memory Policy: Making Memory Utility Identifiable by Intervening on Retrieval
Arman Behnam and Binghui Wang
Memory-augmented large
language models must decide which
memories to retain, and recent systems do so by estimating each
memory's effect on task performance. However, these estimates rely entirely on retrieved
memories. When a
memory is never retrieved, store-level interventions produce identical outcomes, leaving its utility unidentified. This is a
retrieval-level positivity violation, invisible to diagnostics that examine only
memory operations. We introduce Causal
Memory Policy (CMP), a causal framework that restores identification by intervening on
retrieval itself, reserving a fixed number of context slots for
memories sampled with known propensities. CMP estimates
memory utility by self-normalized inverse propensity weighting under a balanced assignment design. We prove the causal factorization of
memory utility through
retrieval, the unbiasedness and exact variance of the
estimator, and the optimal decision rule under irreversible operations. Empirically, identification fails for 54% of required
memories on LongMemEval and 67% on LoCoMo, and the failure persists in a deployed
memory system. CMP improves discrimination between required and non-required
memories from 0.54 to 0.66 AUC. Finally, we show that identified
memory utility alone is insufficient for retention decisions: per-query utility reaches 0.78 AUC on the query for which it is estimated, yet no aggregation available to a retention policy predicts a
memory's value on unseen queries. Code is available at: https://anonymous.4open.science/r/cmp-release-D0C3/.
Read the full abstract ▾
Machine Learning · Cross-listed from physics.ao-ph
AI Emulation of Stochastic Sudden Stratospheric Warming with Interpretable Latent Structure
C. Daniel Boscu, Daniel Hernandez, Fabio Alvarez Ventura et al. (+4 more)
Rare weather regime transitions pose a challenge for data-driven modeling due to class imbalance. In this study, we develop a probabilistic deep learning emulator for a prototypical system with regime transitions, the
stochastic Holton–Mass model of stratospheric variability, and analyze the structure of its learned latent space. The Holton–Mass model exhibits two metastable regimes, a strong and a weak polar
vortex, maintained by nonlinear wave–mean flow interactions, with weak
stochastic forcing intermittently triggering rare transitions between these regimes that qualitatively represent SSW events. We employ a ResNet-inspired Conditional Variational Autoencoder with six-layer encoder and decoder layers and explicit current-state conditioning to model the distribution of the system's state at the next time step (one day). The emulator accurately reproduces short-term dynamics, steady-state probability distributions, regime persistence statistics, rare transition rates, the transition committor function, and the transition expected lead time of the physical model. Beyond emulation fidelity, we interrogate the learned latent representation to understand how the model internalizes the underlying metastable structure of the dynamics. Principal Component Analysis of the 32-dimensional latent space reveals a clear and unsupervised separation into four physically interpretable clusters corresponding to strong versus weak
vortex regimes and stable versus transition-prone configurations. Such emergent regime separation in latent space is hard to identify for deep generative models applied to high-dimensional
stochastic systems. Our results show that carefully designed probabilistic emulators can uncover physically meaningful
manifolds governing extreme-event dynamics, potentially aiding the development of improved operational advanced warning systems.
Read the full abstract ▾
All authors C. Daniel Boscu, Daniel Hernandez, Fabio Alvarez Ventura, Justin Finkel, Ashesh Chattopadhyay, Pedram Hassanzadeh, Dorian S. Abbot
Machine Learning · Cross-listed from quant-ph
Sequential Capacity of Quantum Processes with Finite Memory
Yibin Wang
How complex can the responses of a
quantum device become as it runs longer with a fixed internal
memory? We quantify this complexity through sequential response capacity: how many adaptive testing stages, each using a fresh run, can continue to separate possible processes by a prescribed gap in response probabilities. For fixed system and
memory sizes, we establish a tight law relating this capacity to run length and probability resolution. At fixed resolution, the capacity grows on the order of $K\log K$, where $K$ is the number of time steps in each run. Our construction attains this growth using time-dependent phase rotations on a single visible
qubit with no additional internal
memory; its tests give response probabilities exactly zero or one. Under the same tests, classical
stochastic processes that measure in a fixed basis at every step have only linear capacity at fixed sizes and resolution. For phase sequences selected by a stored classical label, we then quantify how known independent Pauli noise changes this logarithmic enhancement. With ideal controls and weak residual phase noise after correction, we prove matching capacity bounds at a fixed small probability gap. These bounds identify the inverse residual phase-flip probability as the coherence timescale that limits the extra logarithmic growth.
Read the full abstract ▾
Machine Learning
Learn the Directions, Normalize the Gains: Post-Training Normalization for LoRA
Zailong Tian, Yanzhe Chen, Zhuoheng Han et al. (+2 more)
While Low-Rank Adaptation (LoRA) enables efficient task specialization, its learned updates can compromise capabilities beyond the target task. We identify adaptation imbalance: a few singular directions dominate the trained update, leaving its performance sensitive to how gains are allocated. We argue that learning where to adapt does not ensure that adaptation gains are well balanced. This motivates LoRA-Norm, a post-training normalization method that retains learned directions while rebalancing their gains. LoRA-Norm combines spectral rebalancing, a fixed nonlinear transformation of singular values, with
nuclear-norm restoration, which preserves the original total spectral mass. It requires no calibration data or additional training and introduces no inference overhead. Across two backbones and three adaptation tasks, LoRA-Norm improves average specialization and capability retention, outperforming the evaluated post-hoc spectral
pruning and gradient-guided editing configurations on both measures. Stronger functional equalization brings no consistent additional gains, revealing that balancing adapter gains and equalizing their responses are distinct objectives.
Read the full abstract ▾
All authors Zailong Tian, Yanzhe Chen, Zhuoheng Han, Houfeng Wang, Lizi Liao
Artificial Intelligence
External Observers May See More Clearly: Cross-Model Span-Level Hallucination Detection in Large Language Models via Hidden State Probing
Kingshuk Gupta and Davide Buscaldi
As Large
Language Models (LLMs) increasingly serve as foundational
reasoning engines, their tendency to hallucinate remains a critical vulnerability. While recent internal state probes offer a promising alternative to slow external
retrieval systems, they largely reduce
hallucination detection to a
token-wise binary classification task, failing to capture the structured, sequential boundaries of semantic drift. Here, we introduce an internal hidden state framework for fine-grained, span-level
hallucination detection. By inspecting layer-wise activation patterns, we attempt to detect the exact
hallucination onset and continuation
tokens in an LLM generation. Our experiments show that this approach successfully isolates
hallucination onsets, achieving substantial improvements in Precision-Recall AUC over random baselines despite extreme class imbalance. Ultimately, we propose a novel cross-model detection framework in which one model observes the internal representations elicited by another model's generation. We find that an external observer can match or exceed a generator's self-detection of its own
hallucination onsets, including when the observer is the smaller model, suggesting that self-detection is not the ceiling for onset localisation.
Read the full abstract ▾
Numerical Analysis
Entropy dissipative high order schemes for a hyperbolic model of two-layer thin film flow
Rahul Barthwal, Philipp Öffner, Chetan Singh et al. (+1 more)
In this article, we develop high-order,
entropy-dissipative schemes for a
hyperbolic system of conservation laws describing the first-order dynamics of two-layer thin-film flows of immiscible fluids with perfectly soluble solute particles. The key aspect is to construct
entropy-conservative fluxes in the sense of Tadmor. Here, we consider and compare two different construction processes: an exact formulation and a commonly used simplified approximation in the flux evaluation. While the approximate approach is simpler at the level of construction and is frequently employed in practice (also in other contexts), it is less structured, leading to increased computational cost in the implementation compared to the exact formulation, which has been much more complicated to derive. We employ these fluxes within both an
entropy-dissipative finite-difference framework and an
entropy-dissipative discontinuous Galerkin spectral element method. Through numerical experiments using two distinct spatial discretizations, we investigate the performance and robustness of the proposed methods and demonstrate that the choice of numerical flux has a noticeable impact on the efficiency of the resulting solvers.
Read the full abstract ▾
All authors Rahul Barthwal, Philipp Öffner, Chetan Singh, Jan Zawallich
Machine Learning
Foundations without Fundamentals: Zero-Shot Blind Spots in Time Series FMs
Nafiseh Ghoroghchian, Haipeng Zhang, Shuyi Han et al. (+2 more)
Despite the success of Time Series Foundation Models (TSFMs) on broad
benchmarks, their ability to internalize basic temporal logic, especially in settings supported by exogenous
covariates, remains under-examined. We introduce SimpleTimeBench, a diagnostic univariate and multivariate "unit test" suite for primitives such as monotonic trends, periodic signals and leading indicator
covariates, scenarios where near-perfect forecasts should be trivial. Surprisingly, prominent multivariate TSFMs (Chronos-2, Moirai and Toto) frequently produce suboptimal zero-shot forecasts for these inputs. While fine-tuning Chronos-2 improves its behaviour on specific tasks, we show that this adaptation degrades performance on other fundamental patterns rather than enhancing its generalizable foundational capabilities. This reveals a gap between pre-training scale and basic temporal
reasoning, suggesting that current TSFMs could potentially lack the inductive biases needed to capture simple predictable functions. We further demonstrate that these failures are not merely synthetic curiosities: they persist in real-world sensor
forecasting, where TSFMs consistently underutilize leading indicators available in observed
covariates. This inability to capture simple relationships limits the practical utility and reliability of current multivariate models.
Read the full abstract ▾
All authors Nafiseh Ghoroghchian, Haipeng Zhang, Shuyi Han, Alex Labach, George Stein
Information Retrieval
Optimizing Effective Training Time for Large-Scale Recommendation Systems
Mingming Ding, Ruilin Chen, Yuzhen Huang et al. (+29 more)
Lifecycle overhead silently consumes accelerator capacity across large-scale recommendation training fleets. Our largest recommendation workloads process tens of billions train- ing examples per day on thousands of GPUs. Before this work, only 50-60% of their end-to-end wall time advanced training on new data. We present a fleet-scale study of this lifecycle overhead and a set of optimizations spanning the full training stack. We use Effective Training Time (ETT%) as an operational framework to instrument lost time, localize it to independently owned infrastructure components, and expose work repeated across job restarts. This analysis guides optimizations like communication elimination and pipeline overlap during trainer initialization; dynamic-shape handling, autotuning
pruning, and reusable Py- Torch 2 compilation caches; asynchronous checkpointing; stan- dalone model publishing; and reductions in recovery cost. We evaluate the optimizations on representative models and measure their impacts in our training fleet. ETT% improves on every
benchmark, by 15.5% on average, and reaches 85% on our largest workload. Fleet-wide ETT% rose from about 80% to above 90% after deployment.
Read the full abstract ▾
All authors Mingming Ding, Ruilin Chen, Yuzhen Huang, Hang Qi, Menglu Yu, San Tan, Damian Reeves, Boris Sarana, Kevin Tang, Satendra Gera, Gagan Jain, Sahil Shah, Vishwa Karia, Fuzail Khan, Yashasvi Makin, Edward Z. Yang, Oguz Ulgen, Jia Chen Ren, Laith Sakka, Mayank Garg, Meet Vadakkanchery, Aici Lin, Wei Sun, Mengjiao Zhou, Shuai Yang, Junqing Zhou, Max Leung, Apoorv Purwar, Musharaf Sultan, John Bocharov, Zhenyu Tang, Vivek Trehan
Systems and Control
Local Consistency Does Not Guarantee Global Conservation: Auditing Zero-Shot Composition of Airway Flow Operators
Nichula Sathmith Wasalathilaka, Navodya Heshan Samarasinghe Dhanujaya Suraweera, Kevin Dawson et al. (+2 more)
Neural operators approximate PDE solutions within a geometry family, but independently learned local operators need not form a consistent global simulator. We study frozen, single-pass composition for steady incompressible flow in idealized two-dimensional airway trees. Separate Tube, bifurcation, and trifurcation DeepONets are trained on 4,872 primitive CFD cases using field supervision and auxiliary divergence, port-flux, component-balance, and port-pressure penalties. The validation-selected deployment is frozen before whole-tree CFD fields are inspected and assembled without tree training, iterative coupling, flux correction, or CFD-informed adjustment. It retains major flow patterns and controlled pathology responses with 0.204-0.215 s CPU inference, but has a 22.68% prescribed-inlet-normalized external residual. A post-hoc sensitivity protocol, frozen before new training and evaluation, repeats Data, Div, and Full models across three seeds with Tube fixed. Relative to Data, Full reduces primitive composite scores by 26.7% for Y2 and 26.4% for Y3 and reduces assembled component-residual and interface-mismatch RMS by 7.2% and 16.6%, respectively. Nevertheless, mean tree velocity error increases by 7.5%, while external residual increases from 17.49 +/- 4.38% to 26.77 +/- 3.65%. Local regularization can therefore improve primitive and assembled local diagnostics without ensuring accurate global fields or conservation.
Read the full abstract ▾
All authors Nichula Sathmith Wasalathilaka, Navodya Heshan Samarasinghe Dhanujaya Suraweera, Kevin Dawson, Chinthaka Jacob Mervyn Parakrama Bandara Ekanayake, Roshan Godaliyadda
Systems and Control
A Dynamic Generalized Kalman Consensus Filter for Switching Sensor Networks
Tirthankar Chakraborty and Arnab Maity
Distributed state estimation is critical for applications such as surveillance, autonomous
navigation, and wide-area monitoring, where sensor
agents must cooperatively track targets using only local measurements and neighbor-to-neighbor communication. Existing distributed filters have been shown to achieve accurate estimation even under sparse inter-
agent communication and limited sensing ranges. However, many of these methods rely on consensus parameters that depend on global properties of the communication graph, such as the maximum degree of the graph, and are therefore sensitive to changes in network
topology. This limitation is particularly significant in sensor networks with mobile
agents, where communication links change over time. This paper presents a Dynamic Generalized Kalman Consensus Filter for target tracking in sensor networks with switching communication
topologies. The proposed
algorithm computes information-based consensus weights using only locally available quantities, eliminating the need for global network parameters. Numerical simulations demonstrate that the proposed
algorithm maintains estimation accuracy under switching network
topologies and outperforms existing distributed filters in the given tracking problem.
Read the full abstract ▾
Robotics
UniWAM: Unified World-Action Model
Wenxuan Song, Jiayi Chen, Jingbo Wang et al. (+13 more)
Vision-
language-action models benefit from the understanding and
reasoning capabilities of pretrained
vision-
language models, but action-only supervision provides limited grounding in world dynamics. Conversely, world-action models inherit spatiotemporal priors from
video generation models, yet remain limited in semantic understanding and
reasoning under distribution shifts. We introduce UniWAM, a unified architecture that integrates a physical reasoner, a world generator, and an action predictor to jointly learn semantic understanding of the physical world, visual generation, and action prediction. To ensure the quality of the training data, we developed a rigorous data cleaning and annotation pipeline for both human
egocentric data and
robot data. To adapt the
vision-
language component to
embodied tasks while preserving its inherited
language capabilities, we represent low-level actions in natural
language and introduce a pre-training recipe that assigns complementary supervision from visual question answering (VQA) data, human
egocentric data, and
robot demonstrations to the appropriate model components. During post-training, future visual noise augmentation reduces reliance on precise future predictions, while history-conditioned flow matching uses encoded action history to initialize action generation. Together, these designs significantly reduce denoising steps while maintaining performance. UniWAM achieves state-of-the-art (SOTA) performance across multiple evaluations, including in-distribution performance, robustness, generalization, instruction following, and long-horizon task execution. Furthermore, we uncover a log-linear scaling law of unified human-
robot co-training, demonstrating the effectiveness of large-scale pre-training on a mixture of human and
robot data.
Read the full abstract ▾
All authors Wenxuan Song, Jiayi Chen, Jingbo Wang, Shuai Zhou, Xicheng Gong, Zehua Fan, Ziyang Zhou, Junwu E, Haodong Yan, Fuhao Li, Qize Yu, Xu Huang, Pengwei Wang, Wen Chen, Shunbo Zhou, Haoang Li
Computer Vision
Learning from Failure: Leveraging Unreliable Predictions in Semi-Supervised Real-World Adverse Weather Removal
Cap Dang Xuan Kiet and Tat-Jen Cham
Adverse weather image restoration aims to recover images degraded by rain, haze, snow, and other weather-induced artifacts, thereby improving the robustness of outdoor
vision systems. Existing unified restoration models exhibit limited generalization to real-world scenes due to their reliance on synthetic supervision and insufficient semantic constraints. In this paper, we propose a novel student–teacher semi-supervised framework that addresses both challenges. Specifically, we introduce an unreliable database that preserves failed teacher predictions as informative negative samples for contrastive learning, while a reliable database stores high-quality teacher predictions as positive samples. By jointly exploiting reliable pseudo-ground truths and unreliable teacher outputs, the proposed framework learns to enhance desirable restoration characteristics while avoiding common failures. We further propose a phase spectrum-based semantic constraint that replaces computationally expensive text-based supervision with an efficient and naturally aligned semantic prior. An adaptive phase consistency loss is also designed to dynamically balance supervision between the degraded input and teacher pseudo-ground truths according to degradation severity. Extensive experiments on real-world
benchmarks demonstrate that the proposed method consistently outperforms existing state-of-the-art approaches in restoration quality and perceptual fidelity while exhibiting stronger generalization to real-world adverse weather conditions.
Read the full abstract ▾
Artificial Intelligence
HydroJEV: A one-second, training-free screen for cyber-attack and fault attribution in water distribution networks
Tianwei Mu, Shengyan Jiang, Mingzhe Yuan et al. (+5 more)
When a SCADA alarm is raised in a water distribution network, operators must decide quickly whether it reflects a cyberattack, a physical fault, a normal transient or a faulty sensor. Supervised classifiers need labelled incidents that utilities rarely have, and frontier large
language models (LLMs) take tens of seconds per decision. We tested whether Jev, a training-free model that returns class probabilities in about one second, can serve as the first tier of this triage. On a four-class cause-attribution
benchmark built on the C-Town network in EPANET, Jev was compared with a hand-written rule tree, a supervised classifier and seven cloud LLMs on identical evidence in four sealed, pre-registered rounds. With only a label-free prior correction, Jev matched the rule tree (macro-F1 0.62-0.64 against 0.56-0.61 in distribution) and exceeded the supervised classifier by 0.36-0.42 on event subtypes absent from its labels, in all four rounds, and it outperformed the classifier whenever fewer than about four labelled events per class were available. Jev also decided 20-40 times faster than frontier LLMs. Accepting only benign Jev verdicts confirmed by the rule tree spared an LLM reviewer 35-38% of windows on fresh sealed sets without loss of macro-F1. Transferred unchanged to two further networks, this gated cascade stayed within the non-inferiority margin of its reviewer on all four sets. A fast, training-free screen can therefore take over about a third of the review load in SCADA anomaly triage while preserving the accuracy of deliberate review.
Read the full abstract ▾
All authors Tianwei Mu, Shengyan Jiang, Mingzhe Yuan, Qing Luo, Min Xiao, Wenhong Wang, Jun Li, Manhong Huang
Computational Complexity
Short Resolution Refutations for CNFs with Bounded Weighted Incidence Treewidth
Shaowei Cai and Ziqun Li
It is an open problem in proof complexity whether every unsatisfiable CNF formula has an FPT-sized resolution refutation parameterized by incidence treewidth. In this paper, we establish several upper bounds on resolution refutation length related to this problem. Consider an unsatisfiable CNF formula $F$ with $n$ variables, $m$ clauses, maximum clause width $k$, and incidence treewidth $\mathrm{tw}^*(F)$. In this paper, we introduce two variants of incidence treewidth. Their definitions can be stated informally as follows. The first is log-weighted incidence treewidth $\mathrm{tw}_{\log}^*(F)$, which is the treewidth of the weighted incidence graph, in which variables have weight one and each clause has weight equal to the logarithm of its width. The second is partially log-weighted incidence treewidth $\mathrm{tw}^*_{\mathrm{plog}}(F)$, which is a refinement of log-weighted incidence treewidth. In this variant, for a nice tree decomposition of the incidence graph, each clause has weight one along a path selected for that clause and elsewhere has weight equal to the logarithm of one plus the number of its literals whose variables do not appear in any bag on that path, and variables have weight one. For every unsatisfiable CNF formula $F$, we prove the existence of (i) an FPT-sized resolution refutation parameterized by log-weighted incidence treewidth, with width at most $\mathrm{tw}_{\log}^*(F)+k$; (ii) a resolution refutation of length $(n+m)k^{O(\mathrm{tw}^*(F))}$ and width at most $\mathrm{tw}^*(F)+k$; (iii) an FPT-sized resolution refutation parameterized by partially log-weighted incidence treewidth; and (iv) an FPT-sized regular resolution refutation parameterized by log-weighted incidence treewidth. Our main idea is to construct FPT-sized $k$-DNF resolution refutations parameterized by incidence treewidth, and then convert them into resolution refutations.
Read the full abstract ▾
Databases
Prune First, Decide Fast: Scalable Semantic Query Processing with JEVDB
Zhengle Wang, Hanxu Yan, Fuheng Zhao et al. (+1 more)
Semantic database systems extend SQL with foundation-model inference over unstructured data, but current engines rely heavily on autoregressive LLMs for discrete relational decisions, creating high latency and monetary cost. We present JEVDB, a scalable semantic database system that uses fast, typed decision models for semantic filters, joins, classification, and ranking, while selectively escalating uncertain cases to generative LLMs. To reduce semantic-join work, JEVDB combines exact Yannakakis-style semijoin reduction over relational structure with Semantic Bloom Filters (SBFs), which use registered necessary conditions to screen candidates across latent semantic edges. We evaluate JEVDB on SemBench and Shelob, a TPC-DS-derived semantic-join workload. On SemBench, JEVDB-Flash achieves the lowest latency on all 21 evaluated queries and the lowest cost on 19, while maintaining competitive answer quality. On Shelob, where joins scale to 540K candidate pairs, JEVDB completes all queries with 95.7%-97.5% mean F1. SBF screening removes 87.4% of candidate pairs before semantic evaluation, and reusable condition-index scoring further reduces
reasoning-model escalations by 55.2%. An interactive query simulator, source code, and
benchmarks are available at https://jevdb.org.
Read the full abstract ▾
All authors Zhengle Wang, Hanxu Yan, Fuheng Zhao, Chunwei Liu
Computer Vision
Form and Void: Entangled Composition through an Autonomous AI Agent
Shiwen Wang, Jian Yang, Xu Wang et al. (+2 more)
Positive and negative space is a fundamental principle in visual composition, supporting visually coherent forms and layered semantic relationships. Generating such compositions is challenging because it requires coordinated control over two semantic concepts that share a common boundary. Although recent text-to-image models and
multimodal large
language models (MLLMs) have achieved strong performance in image generation and visual understanding, positive-negative space generation remains difficult, particularly under direct single-pass prompting. In this work, we present the Form and Void
Agent (FaV-A), a
multimodal agent designed for staged positive-negative space generation. FaV-A follows a progressive workflow: it first generates a base object, then analyzes its shape and spatial structure to identify candidate negative-space semantics, and finally produces compositional instructions for the final image generation stage. Experimental results and ablation analyses suggest that FaV-A provides a more effective framework than direct zero-shot MLLM baselines for producing visually coherent and semantically aligned positive-negative space compositions.
Read the full abstract ▾
All authors Shiwen Wang, Jian Yang, Xu Wang, Xincan Wang, Weiming Dong
Computer Vision
DiDE:Direct Injection with Color-Texture DEcoupling for 3D Stylization
Tao Wu, Alexandra Gomez-Villa, Senmao Li et al. (+3 more)
Recent advances in rectified flow-based image-to-3D generative models have enabled high-fidelity 3D asset generation. Building on this, a growing line of work has exploited these strong 3D priors for training-free stylization, transferring visual attributes from a reference image onto a generated 3D asset. However, existing methods enforce an all-or-nothing paradigm: color and texture are transferred jointly, with no mechanism to control them independently – a limitation we formalize as Disentangled 3D Stylization(Disen3D). To address this, we propose DiDE, the first training-free framework for Disen3D. Key to our approach is the observation that the structured latent space of image-to-3D models is overcomplete with respect to texture: texture information occupies only a small subset of the style-significant channels, leaving a free subspace available for independent color encoding. DiDE exploits this via a channel partition mechanism that processes a content image, a texture reference, and a color reference through dedicated branches and composes both style signals interference-free at every self-
attention layer, preserving content geometry throughout. Experiments on Disen3D-Bench, our newly collected multi-reference
benchmark, show that DiDE consistently outperforms 2D and 3D stylization baselines in color fidelity, texture transfer, and content preservation.
Read the full abstract ▾
All authors Tao Wu, Alexandra Gomez-Villa, Senmao Li, Yaxing Wang, Joost van de Weijer, Kai Wang
Machine Learning
Distributionally Robust Schrödinger Bridge
Jinhwan Sul, Panagiotis Theodoropoulos, Vincent Pacelli et al. (+2 more)
Schrödinger bridge (SB) learns
stochastic transport between prescribed initial and target distributions. When the initial distribution shifts at test time, the learned dynamics can fail to recover the target distribution. We introduce the Distributionally Robust Schrödinger Bridge (DRSB), which learns a single controller that accounts for uncertainty in the initial distribution. The DRSB objective consists of control energy and a KL penalty between the resulting terminal distribution and the target distribution. DRSB seeks a single controller that minimizes the worst-case value of this objective as the initial distribution varies within an ambiguity set around the nominal distribution. We derive an exact variational formulation of this objective and connect its fixed-terminal-cost subproblem to
stochastic optimal control and distributionally robust optimization. This formulation motivates an alternating
algorithm that updates the
adversarial initial distribution, estimates the terminal log-density ratio, and trains the controller. We develop Wasserstein and Sinkhorn variants using
stochastic control optimality conditions to approximate the gradients required for
adversarial updates. Experiments on two-dimensional transport tasks and image-to-image translation show improved robustness to input perturbations relative to standard SB, with a tradeoff in nominal performance. On Gaussian mixture transport, Sinkhorn DRSB also achieves lower mean sliced Wasserstein distance than fixed-level noise augmentation at both tested unseen noise levels.
Read the full abstract ▾
All authors Jinhwan Sul, Panagiotis Theodoropoulos, Vincent Pacelli, Jaemoo Choi, Evangelos Theodorou
Computation and Language
Typological Alignment of Stack-Based Language Models on Mildly Context-Sensitive Artificial Languages
Nadine El-Naggar, Tatsuki Kuribayashi and Ted Briscoe
Some properties of
languages, e.g., subject-object-verb (SOV) word order, are more prevalent than others among the thousands of attested natural
languages (NLs). Such typological commonality is often attributed to learning biases. Computational simulations, recently with
language models (LMs), have facilitated the exploration of this theory. In this paper, we extend existing analyses of the relationship between LMs' learning biases and typological commonality on both data and model sides, focusing on: (i) cross-serial dependencies, the upper limit of attested syntactic complexity, and (ii) stack-based LMs (SLMs), potentially facilitating learning of hierarchical patterns. We first evaluate generalization of SLMs on cross-serial dependencies across diverse artificial
languages and confirm that they struggle with such constructions. However, SLMs with limited working
memory generalize better suggesting a possible basis for such inductive bias and thus the typological commonality of some word order configurations.
Read the full abstract ▾
Machine Learning
CARM: Cancellation-Aware Response Masking for LLM Reinforcement Learning
Yafei Zhang, Songshuo Lu, Sicong Liao et al. (+2 more)
Recent years have witnessed the rapid adoption of reinforcement learning (RL) in large
language model (LLM) post-training, with substantial gains in mathematical
reasoning and code generation. In practical systems, however, policy updates and differences between rollout and training engines can make sampled responses off-policy. Sequence-level masking addresses this mismatch by deciding whether an entire response should contribute to optimization. A common masking rule uses the length-normalized
geometric mean of sampled
token probability ratios. Its signed log-ratios can cancel across positions, concealing substantial bidirectional policy drift. We propose Cancellation-Aware Response Masking (CARM), a sequence-level mask that takes the absolute value of each
token log-ratio before averaging, preventing opposing probability changes from canceling. We prove that accepted responses satisfy a joint bound on the fraction of sampled-
token ratios outside a prescribed band and their mean log-distance beyond its boundaries. Experiments on mathematical
reasoning and code generation show that CARM improves mean@16 averaged over AIME 2024/2025/2026 and BeyondAIME by up to $3.13$ percentage points over
geometric-mean masking, and increases average pass@1 across four code
benchmarks by $2.88$ points over the strongest evaluated baseline. These findings support CARM as a theoretically grounded and effective method for response-level off-policy control in LLM reinforcement learning.
Read the full abstract ▾
All authors Yafei Zhang, Songshuo Lu, Sicong Liao, Zhi Chen, Yaohua Tang
Artificial Intelligence
Mimir: Physics-Grounded LLM Agents for Long-Horizon Irrigation Control
Yimeng Liu, Mi Zhang, Younsuk Dong et al. (+1 more)
Large
language model (LLM)
agents increasingly combine
reasoning, tool use, and action, but most evidence comes from episodic tasks with relatively immediate feedback and reset failures. Long-running physical control operates in a different regime: actions alter future states, errors compound across decisions, and an
agent must improve from experience without being allowed to rewrite the physical rules that make execution safe. We study this regime through irrigation, where daily decisions interact with soil-water dynamics over entire growing seasons. We present Mimir, a physics-grounded LLM
agent organized around two repair timescales. At the fast timescale, a structured physical interface and deterministic simulator turn an LLM output into a proposal that we numerically check, revise, and subject to bounded deterministic action selection before execution. At the slow timescale, recurrent failure patterns are consolidated into persistent contextual principles that condition future proposals, while the physical model, evaluator, and execution constraints remain immutable. Under a common retrospective evaluator across multiple sites, crops, and years, Mimir attains the lowest reported aggregate control cost among the evaluated references and uses about 51% less irrigation than the historical schedule replay. The ablation study show higher control cost when forward simulation, verified revision, or persistent context is removed; model-scale and model-family studies show no monotonic gain from increasing LLM size. The resulting lesson show that persistent physical
agents can combine semantic
reasoning with bounded, evidence-driven self-improvement while reserving physical truth and actuator authority for explicit numerical mechanisms.
Read the full abstract ▾
All authors Yimeng Liu, Mi Zhang, Younsuk Dong, Zhichao Cao
Artificial Intelligence
Global Coherence: When Every Agent Is Right and the Team Is Still Wrong - A Local-to-Global Semantic Foundation for Multi-Agent Collaboration
Xin Heng
AI
agents can each make locally valid decisions yet jointly produce an invalid result. We call this the global coherence problem: a failure of shared state, not merely of intelligence. Our Observation-Aliasing Impossibility
Theorem gives the exact boundary. A policy can guarantee a valid action exactly when all worlds producing the same observation share an admissible action. If k indistinguishable worlds require pairwise-disjoint actions, the best randomized worst-case success is 1/k; more
reasoning, roles, messages, or samples cannot recover the missing distinction. A stronger model can reason better within its context, but it cannot see beyond it. The smallest sufficient context is NP-hard to find, yet counterexample-guided refinement finds a sufficient one and recertifies it as rules change. We give local-to-global runtime semantics X = (H, C, G, F; D):
topology H records overlapping scopes; category C governs state-changing actions; groupoid G keeps reversible translations; sheaf F tests whether local views glue into one world; and history D keeps only what alters legal futures. Each preserves the needed information as state is split, translated, or changed. Models propose; the harness owns shared state and governs commit. Ten studies test the boundary. On a revision
benchmark, a frontier model scores 40/40 when the deciding event is visible; hidden, every arm is consistent with chance (1/3); one restored fact returns 40/40. On TeamBench, ordinary teams exceed a shared budget in 5/5 runs, a visible live count leaves 4/5, and commit enforcement leaves 0/5. In tau2-bench Telecom, current-state checks score 0.07 after silent reverts; the harness scores 1.00. Refinement keeps a sufficient context through rule changes at 28% of the all-field cost. Solvers that already own the state tie it, as predicted. Local intelligence cannot substitute for missing global state.
Read the full abstract ▾
Machine Learning
Relative Transitions, Not Absolute Destinations: A Transfer-and-Ground Framework for Target-Trajectory-Free Human Mobility Generation
Yidi Wang, Yunhe Zhang, Bangchao Deng et al. (+2 more)
Individual mobility trajectories support urban analysis and location-based services, yet most trajectory generators require observations from their deployment city. This assumption excludes precisely the cities where trajectories are unavailable even though points of interest (POIs) and their attributes can be obtained from public maps. We study target-trajectory-free generation: learning from POIs and trajectories in source cities while utilizing only POI coordinates and categories in a target city, with no target trajectory or trajectory-derived statistic available for training, model selection, or generation. Existing trajectory generators typically predict absolute destinations, entangling reusable movement behavior with city-specific POI identities and spatial layouts. Our core insight is to replace this city-bound output with context-conditioned relative transitions. We propose Nomad, a transfer-and-ground framework that separates learning how people move from determining where those movements are realized. Specifically, a history-conditioned flow-matching model learns from source trajectories a transition prior over semantic displacement between POI contexts, geographic displacement, and elapsed time; at inference, a behavior graph and an exploration–return walk ground sampled transitions onto the target POI map. This factorization enables a direct test of representation level transferability without assuming invariance of the full mobility distribution. Extensive experiments across ten cities and 14 transfers show that Nomad outperforms adaptation baselines in trajectory fidelity and downstream utility, lowering the average error over the best baseline of each metric by about 15% in distributional fidelity and about 3% in downstream utility.
Read the full abstract ▾
All authors Yidi Wang, Yunhe Zhang, Bangchao Deng, Dingqi Yang, Pengyang Wang
Artificial Intelligence
SPHERE: Adaptive VR Indoor Scene Generation via LLM-Enhanced Spatial Preference Learning and Human-in-the-Loop RL
Hyeonmin Lee, Zheng Wei, Kyungmin Kwon et al. (+3 more)
While Large
Language Models (LLMs) advance 3D indoor scene synthesis, current pipelines fail to retain user-specific preferences across sessions, making immersive authoring a repetitive and physically fatiguing process. We present SPHERE, an adaptive VR generation framework that transforms isolated synthesis into continuous human-AI co-creation. SPHERE extracts persistent spatial preferences from natural
multimodal interactions (
speech and controller edits). To ensure
geometric resilience against spatial distortions, it abstracts these raw edits into hierarchical constraints modeling both local functional and global topological contexts. Furthermore, a human-in-the-loop reinforcement learning mechanism dynamically updates
retrieval policies based on the user's final edited scenes. A mixed-design user study ($N=42$) and an offline ablation demonstrate that SPHERE significantly reduces corrective edits and physical demand, preventing bias toward shallow object-level traits to yield geometrically resilient, profile-aligned layouts. Ultimately, SPHERE demonstrates how capturing demonstrated spatial logic enables controlled spatial adaptation, establishing a reliable, governed human-AI collaboration framework for immersive authoring. Project page and source code will be available at: https://github.com/hyeonmin11/SPHERE
Read the full abstract ▾
All authors Hyeonmin Lee, Zheng Wei, Kyungmin Kwon, Jumin Seo, Jiwon Park, Hayoung Oh
Computation and Language
Old Ideas, Novel Problems: The Instability of LLM-Based Novelty Evaluation
Noy Sternlicht, Simra Shahid, Peter Jansen et al. (+3 more)
AI-based ideation systems are often evaluated on the novelty of their ideas, and that judgment is increasingly delegated to large
language models. Such judges are typically built ad hoc and validated, if at all, on human-authored papers rather than on the generated ideas they are meant to score. So, how do novelty judges perform? Not well. We present a systematic controlled study of novelty evaluation design choices. We build an evaluation set automatically, mining OpenReview for passages where reviewers explicitly affirm or dispute a paper's originality and keeping only submissions with unanimous agreement at the extremes of their research area; we pair these with ideas from a vanilla LLM generator. Across six judges, we find that seemingly small
prompt design choices have large consequences; e.g., simply telling the judge that reviewers found one idea novel and the other not can change its verdict on more than half of the identical idea pairs it is shown, shifting pairwise accuracy by over 50 points and occasionally pushing it below chance. The same change helps one judge and hurts another.
Retrieval and larger
reasoning budgets help little, and two purpose-built novelty evaluators are outperformed by our cheapest prompted baseline. These results raise questions about reported novelty gains of automated ideation systems, and call for robust novelty evaluation methods.
Read the full abstract ▾
All authors Noy Sternlicht, Simra Shahid, Peter Jansen, Daniel S. Weld, Pao Siangliulue, Tom Hope
Computer Vision
Task-Adaptive Grounded 3D-Programmers Using 2D VLMs
Arman Raayatsanati, Sombit Dey, Anna-Maria Halacheva et al. (+3 more)
Recent
vision-
language models (VLMs) exhibit remarkable generalization and
reasoning abilities, yet 3D understanding in these models is limited by data scale, training diversity, and
reasoning capacity. Instead of naively extending these models into 3D, we take a different approach: we enable powerful 2D VLMs to operate reliably in 3D by introducing 3D grounding and iterative feedback loops with two novel concepts: Canonical Coordinate Framing (CCF) and Task-Adaptive Feedback (TAF). CCF serves as a unified visual representation that anchors both inputs and outputs to a shared Euclidean coordinate system, solving common challenges in 3D grounding such as axis ambiguity, inconsistent metric scale, and floating references. Complementary to this structured framing of the 3D inputs, TAF closes the
reasoning loop with task-adaptive dynamic feedback that enables 2D VLMs to perform varied open-vocabulary tasks within their native visual context. Building on this foundation, we introduce 3D-Prog, a 3D understanding,
reasoning, and generation framework that jointly employs the capabilities of CCF and TAF together with powerful VLMs. Without requiring any retraining, 3D-Prog performs open-vocabulary 3D understanding, manipulation, and generation across both object-level and scene-level tasks. Our experiments show that the joint use of CCF and TAF transforms 2D VLMs into geometry-aware 3D programmers, achieving consistent, interpretable, and high-quality results across diverse 3D tasks.
Read the full abstract ▾
All authors Arman Raayatsanati, Sombit Dey, Anna-Maria Halacheva, Jan-Nico Zaech, Luc Van Gool, Danda Pani Paudel
Computation and Language
Controllable Multi-label Video Safety Detection via Adaptive Tversky Policy Optimization
Guangyu Yang, Jingbiao Mei, Mingsheng Sun et al. (+5 more)
The rapid growth of
video-based social media has increased users' exposure to harmful content, creating a need for reliable automated
video safety detection. Although recent
Vision-
Language Models (VLMs) show strong
video understanding capabilities, existing harmful
video detection systems face two key limitations: they typically reduce safety detection to binary classification, overlooking the inherently multi-label nature of unsafe
videos, and they rely on static training objectives that do not support controllable precision-recall trade-offs, though the desired operating point may vary across moderation pipelines and unsafe categories. To address these gaps, we propose Adaptive Tversky Policy Optimization (ATPO), a reinforcement learning framework for Multi-label
Video Safety Detection (Multi-VSD). ATPO introduces the Adaptive Tversky Reward (ATR), which dynamically adjusts false-positive and false-negative penalties during training to enable controllable precision-recall trade-offs. Experiments on SafeWatch-Bench and XD-Violence show that ATPO substantially improves multi-label performance, increasing the Jaccard Index from 40.66 to 75.44 on SafeWatch-Bench-Real. Moreover, ATR enables reliable steering of the precision-recall operating point, supporting deployment scenarios with heterogeneous policy requirements. Code and checkpoints are provided at https://bruceyg.github.io/ATPO-project-page/ .
Read the full abstract ▾
All authors Guangyu Yang, Jingbiao Mei, Mingsheng Sun, Jinghong Chen, Yingtong Bu, Pengda Qin, Da Chen, Bill Byrne
Data Structures and Algorithms
Vertex-Failure Distance Oracles and Labeling Schemes: Compact and Constant-Approximate
Yaowei Long
We present new
algorithms for the vertex-failure distance oracles and labeling schemes problems in undirected weighted graphs. A vertex-failure distance oracle is a data structure that, given two vertices $x$ and $y$ and a failed vertex set $F$ of size at most $f$, returns an approximation to the distance between $x$ and $y$ in $G \setminus F$. In the labeling-scheme setting, the data structure needs to be stored distributively as labels on the vertices, and each query $(x,y,F)$ must be answered by accessing only the labels of the vertices in $F \cup \{x,y\}$. For any $f\geq 1$ and $k \ge 1$, we obtain a vertex-failure distance oracle with $O(k^{6})$ approximation, space $\tilde{O}(f^{2}n^{1+1/k})$, query time $\tilde{O}(f^{5}n^{1/k})$, and
polynomial preprocessing time. In particular, this is the first time-efficient oracle for multiple vertex failures with space close to linear, as well as the first constant-approximation oracle with
polynomial space when tolerating $\Omega(\log n)$ vertex failures. The previous results, due to [Duan-Gu-Ren, SODA'21], gave two alternatives: for any constant $c \ge 1$ and $\epsilon>0$, one oracle has $\mathrm{poly}(\log n,f)$ approximation, space $n^{2+1/c}\mathrm{poly}(\log n,f)$, and query time $\mathrm{poly}(\log n,f^{c})$, while the other has $(1+\epsilon)$ approximation, space $n^{2+1/c}(\log n/\epsilon)^{O(f)}$, and query time $\mathrm{poly}(\log n,f^{c},1/\epsilon)$. We also obtain a vertex-failure distance labeling scheme with $O(k^{6})$ approximation and label size $f^{3}n^{1/k}\log^{O(k)} n$. This is the first nontrivial distance labeling scheme for vertex failures. Our techniques build on recent tools related to length-constrained vertex expanders and also introduce a new expander-based shortcut sparsification. The latter also leads to a deterministic vertex-failure connectivity labeling scheme of size $\tilde{O}(f^{2})$.
Read the full abstract ▾
Machine Learning
On Language Drift during RLVR Post-Training
Michael Sullivan and Alexander Koller
Recent advances in LLM
reasoning models—driven primarily by the paradigm of post-training via reinforcement learning with verifiable reward (RLVR)—have enabled them to accomplish impressively complex tasks. However, in parallel with their rising capabilities, LLMs have increasingly displayed signs of
language drift in their chains of thought (CoTs): unusual, non-standard, and seemingly nonsensical
language use. Although it is well-documented—and can potentially impair CoT monitorability—the causes of
language drift are thus far poorly understood. In this paper, we identify the conditions under which
language drift occurs: we prove theoretically that RLVR optimization pressure permits unbounded
language drift, while supervised fine-tuning does not. We then show empirically that
language drift specifically arises during RLVR on novel
reasoning tasks—i.e. when the target behavior cannot be drawn out of the base model. Finally, we prove that it is not possible to constrain
language drift without constraining expected reward, suggesting that CoT monitorability cannot be improved without harming performance during RLVR post-training at the frontier.
Read the full abstract ▾
Artificial Intelligence
Atoms to Processes: The Role of Artificial Intelligence and Machine Learning in Chemical Engineering
Michael Baldea, Linda J. Broadbelt, Marianthi G. Ierapetritou et al. (+17 more)
The rapid maturation of artificial intelligence (AI) and machine learning (ML) has catalyzed a profound shift in how chemical engineering problems are formulated, analyzed, and solved. Advances in computing, data availability, and learning
algorithms have enabled AI/ML methods to impact applications spanning atomic-scale simulations, materials and catalyst discovery, transport and thermodynamics, separations, process systems engineering, and industrial operations. This article provides a perspective on recent methodological developments and representative applications, emphasizing how AI/ML tools are being integrated with first-principles models to address challenges of predictive accuracy, data scarcity, extrapolation, interpretability, and model lifecycle management. Across domains, a unifying trend is the move away from purely black-box approaches toward hybrid and physics-informed frameworks that explicitly respect conservation laws, thermodynamic consistency, and known structural constraints. These approaches not only improve robustness and reliability, but also enable meaningful human-AI collaboration by providing information at an appropriate level of abstraction for the task and decision context. We conclude that AI and ML are not replacing the core principles of chemical engineering; rather, they are amplifying them. As the field advances toward increasingly autonomous, adaptive, and sustainable systems, the thoughtful integration of AI/ML with first-principles understanding and domain expertise will be essential to realizing their full potential across both research and industrial practice.
Read the full abstract ▾
All authors Michael Baldea, Linda J. Broadbelt, Marianthi G. Ierapetritou, Akhilesh Jain, Ankur Kumar, Thomas A. Kwan, Fèlix Llovell, Andrew J. Medford, Ilias Mitrai, Joel Paulson, Junyi Qiao, Matthew P. Rivera, Kirti C. Sahu, Lev Sarkisov, Zachary P. Smith, Calvin Tsay, Ching-Mei Wen, Victor M. Zavala, Huacheng Zhang, Dan Zhao