arXiv:2605.31139v1 Announce Type: cross
Abstract: Quantum computers hold great promise for molecular simulation, but noise remains a fundamental obstacle. We introduce a Quantum Monte Carlo (QMC) pre-screening procedure that constructs compact, physically motivated Givens rotation ans\"atze tailored to realistic quantum hardware. By identifying the most important wavefunction contributions early in a QMC simulation, we build circuits that are shallower that conventional alternatives while preserving number symmetry. Benchmarked on Quantinuum System Model H1, QMC-prescreened circuits outperform more complex ans\"atze under realistic noise conditions. The method offers a practical path toward chemical accuracy on quantum devices, by providing an adjustable trade-off between expressivity and circuit depth to generate shallow circuits suited to current high-noise devices, as well as deeper, more expressive circuits that can be deployed on future lower-noise devices.
Science Journals
arXiv:2605.30792v1 Announce Type: cross
Abstract: Speech translation systems increasingly span speech-to-text translation (S2TT), speech-to-speech translation (S2ST), offline translation, and streaming generation, producing outputs that differ in modality, speech realization, and timing behavior. Existing evaluation practices assess important aspects such as translation quality, speech quality, and temporal quality, but these aspects are often evaluated under separate protocols, making it difficult to compare heterogeneous systems comprehensively. To address this gap, we present OpenSTBench, a unified multidimensional evaluation framework that organizes heterogeneous speech translation outputs into a shared evaluation format. OpenSTBench supports both S2TT and S2ST systems in offline and streaming settings, and jointly evaluates translation quality, speech quality, speaker preservation, emotion and paralinguistic fidelity, temporal consistency, and latency. Through experiments on representative speech translation systems, we show that systems with strong translation quality can still differ substantially in speech quality, as well as in temporal quality. OpenSTBench provides a reproducible protocol for analyzing these cross-dimensional differences and supporting application-oriented comparison of speech translation systems. The code and datasets are available at https://github.com/sjtuayj/OpenSTBench.
arXiv:2605.31462v1 Announce Type: cross
Abstract: The stretching behavior of linear macromolecules undergoing conformational transitions is investigated. An exact solution is provided for a two-state system within the elastic freely jointed chain model. This minimal framework contains the smallest set of parameters required to describe such transitions: two Kuhn lengths, two elastic force constants, a free energy difference between both states and a nearest-neighbor interaction energy accounting for cooperativity. Explicit analytical expressions are derived for the chain extension and the probabilities of each state as functions of the applied force.The approach accurately reproduces the experimental force-extension curves of poly(ethylene-glycol) (PEG) and hyaluronic acid (HA), revealing no cooperativity for PEG and negative cooperativity for HA. It also describes the B-DNA to S-DNA conformational transition, a process that exhibits positive cooperativity.We analyze the mathematical conditions required for a transition and identify two fundamental driving mechanisms: differences in Kuhn lengths and differences in force constants.Extensions of the model to systems with more than two conformational states per Kuhn segment are also discussed. The results presented here apply equally to transitions that are intrinsic to the macromolecular structure or induced by ligand-receptor interactions, unifying both cases within a single thermodynamically consistent framework.
arXiv:2603.09787v2 Announce Type: replace
Abstract: Explainable artificial intelligence (XAI) aims to provide human-interpretable insights into the behavior of deep neural networks (DNNs), typically by estimating a simplified causal structure of the model. In existing work, this causal structure often includes relationships where the presence of a concept is associated with a strong activation of a neuron. For example, attribution methods primarily identify input pixels that contribute most to a prediction, and feature visualization methods reveal inputs that cause high activation of a target neuron - the former implicitly assuming that the relevant information resides in the input, and the latter that neurons encode the presence of concepts. However, a largely overlooked type of causal relationship is that of encoded absences, where the absence of a concept increases neural activation. In this work, we show that such missing but relevant concepts are common and that mainstream XAI methods struggle to reveal them when applied in their standard form. To address this, we propose two simple extensions to attribution and feature visualization techniques that uncover encoded absences. Across experiments, we show how mainstream XAI methods can be used to reveal and explain encoded absences, how ImageNet models exploit them, and that debiasing can be improved when considering them.
arXiv:2605.21168v3 Announce Type: replace
Abstract: Safety-critical scenarios are central to evaluating autonomous driving systems, yet their rarity in naturalistic logs makes simulation-based stress testing indispensable. Most scenario generation methods treat surrounding agents as adversaries, but they either (i) induce failures without explicitly modeling vehicle-road physical limits, yielding visually extreme yet physically unsolvable crashes, or (ii) enforce physical feasibility or policy feasibility in isolation, which can over-focus on aggressive maneuvers or remain tied to a controller-dependent capability boundary. We propose ScenePilot, a feasibility-guided, boundary-driven framework that targets the boundary band: scenarios that are physically solvable in principle yet still cause the deployed autonomy stack to fail. We formulate generation as constrained multi-objective reinforcement learning, combining an RSS-derived physical-feasibility score $\sigma$ with an online-learned AV-risk predictor $\Phi$, and introduce step-level feasibility-aware shielding to keep exploration near the feasibility boundary while avoiding infeasible artifacts. Experiments on SafeBench with multiple planners show that ScenePilot yields substantially higher collision rates (+6.2 percentage points) while preserving physical validity, and that adversarial fine-tuning on these boundary-band scenarios consistently reduces downstream crash rates. The code is available at https://github.com/QiyuRuan/ScenePilot.
arXiv:2605.30914v1 Announce Type: new
Abstract: Automated formal verification remains challenging for large language models because data for proof assistants and verification-aware languages is scarce, and correctness depends on satisfying precise machine-checkable specifications rather than producing plausible code. This thesis studies how verifier environments can improve LLM generation of verified programs and proofs through reinforcement learning from verifiable rewards (RLVR) and verifier-guided inference-time search. First, we train open-source models in Dafny with RLVR using Group Relative Policy Optimization (GRPO) and related variants, assembling generated candidates into complete programs and scoring them with compiler and verifier outcomes. Initial experiments on an APPS-derived Dafny dataset increased verified reward from 2.2% to 58.1%, but revealed specification hacking, where models exploit weak formal specifications instead of implementing the intended solutions. After filtering underspecified and vulnerable tasks, multi-turn RLVR on the refined benchmark improves the verified pass rate from 9.7% to 31.1%. Second, we develop a verifier-guided inference scaffold in Lean that treats proof generation as structured search over decomposed subgoals, verifier feedback, diagnostics, and repair. With a fixed base model, the full scaffold with proof reviser improves pass rate on an initial VeriCoding pilot set from 46.2% under direct repair to 69.2%. On the larger VERINA dataset, whole-task decomposition plus proof reviser solves 7 of 42 previously unsolved tasks. We also introduce Dalek-Bench, a repository-scale Lean benchmark derived from the Rust $\texttt{curve25519-dalek}$ verification project; preliminary results remain weak, indicating that stronger progress evaluation and task-specific tool-use policies are still needed.
arXiv:2605.31011v1 Announce Type: new
Abstract: The Saint-Venant shallow water equations (SWE) govern depth-integrated free-surface flows arising in dam-break inundation, flood routing, tsunami runup, and estuarine tidal dynamics. Closed-form analytical solutions exist only for highly idealized Riemann configurations, making rigorously verified numerical solvers essential. This work presents amerta, an open-source Python library that solves the one-dimensional frictionless Saint-Venant system on a uniform Cartesian grid using Monotone Upstream-centered Schemes for Conservation Laws (MUSCL) reconstruction with a minmod slope limiter, the Harten-Lax-van Leer-Contact (HLLC) approximate Riemann solver, and two-stage strong-stability-preserving Runge-Kutta (SSP-RK) time integration. Numba just-in-time (JIT) compilation accelerates the performance-critical kernels. The solver is verified end-to-end against the four canonical Riemann configurations: wet-bed dam break, dry-bed dam break, double rarefaction, and double shock. A six-component post-processing pipeline quantifies space-time topology, final-time error norms with empirical quantile decomposition, self-similarity collapse onto the analytical Riemann fan, integral-norm evolution, boundary-flux-corrected mass and energy diagnostics, and phase-plane analysis against analytical wave curves. The implementation conserves discrete mass to floating-point precision, satisfies discrete entropy admissibility identically, and reproduces all four analytical wave-curve geometries to within sub-centimetre accuracy in the depth-velocity phase plane. The complete source code, analytical-solution evaluators, post-processing scripts, and Network Common Data Format (NetCDF) archives are released under the MIT license.
arXiv:2603.18382v2 Announce Type: replace
Abstract: Anonymization is often assumed to protect privacy once explicit identifiers are removed, because re-identification has historically required specialized expertise, tailored algorithms, and manual corroboration. We show that LLM-based agents weaken this barrier: by combining scattered, individually non-identifying cues with public evidence, they reconstruct real-world identities, sometimes even during benign tasks. We evaluate this risk across three settings -- classical linkage incidents, a controlled benchmark (\emph{InferLink}) that varies fingerprint type, task framing, and attacker knowledge, and open-ended human--AI interaction traces. In the sparsest regime of the Netflix Prize deanonymization setting, agents reconstruct 79.2\% of identities, against 56.0\% for a classical matching baseline; on \emph{InferLink}, they link individuals even without an explicit re-identification request, and more often once one is given. In redacted human--AI interaction traces, agents further resolve anonymized profiles to specific individuals by corroborating contextual cues with public evidence. These findings suggest that privacy evaluations for agentic systems should measure not only what information is accessed or disclosed, but also what identities can be inferred.
arXiv:2603.13727v2 Announce Type: replace
Abstract: Symbolic regression is a powerful tool for knowledge discovery, enabling the extraction of interpretable mathematical expressions directly from data. However, conventional symbolic discovery typically follows an end-to-end, "one-step" process, which often generates lengthy and physically meaningless expressions when dealing with real physical systems, leading to poor model generalization. This limitation fundamentally stems from its deviation from the basic path of scientific discovery: physical laws do not exist in a single form but follow a hierarchical and progressive pattern from simplicity to complexity. Motivated by this principle, we propose Chain of Symbolic Regression (CoSR), a novel framework that models the discovery of physical laws as a chain of symbolic knowledge. This knowledge chain is formed by progressively combining multiple knowledge units with clear physical meanings along a specific logic, ultimately enabling the precise discovery of the underlying physical laws from data. CoSR fully recapitulates the progressive discovery path from Kepler's third law to the law of universal gravitation in classical mechanics, and is applied to three types of problems: turbulent Rayleigh-Benard convection, viscous flows in a circular pipe, and laser-metal interaction, demonstrating its ability to improve classical scaling theories. Finally, CoSR showcases its capability to discover new knowledge in the complex engineering problem of aerodynamic coefficients scaling for different aircraft.
arXiv:2603.11600v2 Announce Type: replace
Abstract: Deep reinforcement learning for continuous control often suffers from high variance, low energy efficiency, and poor generalization under distribution shift, as purely data-driven exploration ignores available physical structure. This paper proposes Hybrid Energy-Aware Reward Shaping (H-EARS), which encodes dominant energy terms -- assumed known a priori -- directly as reward potentials at O(n) per-step computation. H-EARS decomposes the shaping potential into task-oriented and energy-based components, supplemented by an action regularization term that deliberately modifies the optimization objective to enforce energy-efficient control. A complete theoretical foundation is established: functional independence of shaping and regularization, energy-based gradient enrichment under positive-definite Hessian conditions, convergence guarantees under function approximation, and approximate potential error bounds. Across four continuous control benchmarks and four baseline algorithms, H-EARS achieves consistent gains in convergence speed, policy stability, and final performance. High-fidelity vehicle simulations validate applicability in safety-critical settings under extreme road conditions.
arXiv:2604.13624v2 Announce Type: replace
Abstract: InGaZnO (IGZO)-channel FeFETs have attracted notable interest thanks to recent advances in endurance, opening up their application space for read-dominated AI memory tiers. This work evaluates the viability of NOR-type IGZO FeFETs for 3D heterogeneous AI memories from a read-centric design-technology co-optimization (DTCO) perspective, spanning on-chip back-end-of-line (BEOL) RAMs and hybrid-bonded memory chiplets, and off-chip, monolithically integrated 3D FeNOR storage-class memories (SCMs). For on-chip BEOL RAMs and memory chiplets, we demonstrate the cross-node bitcell footprint scalability of IGZO FeFETs capable of delivering down to 10-A SRAM-equivalent bitcell area ($\sim$0.016 $\mu$m$^2$) with 7-nm ground rules while maintaining a sub-5 ns random access latency -- despite their writability challenges. We further identify the sensing margin penalty in NOR FeFET arrays arising from sneak current associated with the negative program-state $V_t$, which requires positive-$V_t$ engineering in order to eliminate the unwanted negative voltage read inhibition -- for example, by ferroelectric layer thinning. Last but not least, we elucidate the read margin implications on 3D FeNOR for SCMs, with the 3D stacking density limited by additional sneak current from neighbor channel shunting.
arXiv:2604.15959v2 Announce Type: replace
Abstract: Multi-objective Bayesian optimization (MOBO) provides a principled framework for optimizing multiple expensive black-box functions. However, existing MOBO methods often struggle with coverage, scalability, and handling constraints and preferences. In this work we propose STAGE-BO, Sequential Targeting Adaptive Gap-Filling $\varepsilon$-Constraint Bayesian Optimization: by analyzing the coverage of the surrogate Pareto front, our method identifies the Pareto front point with the largest uncovered gap, and uses its coordinates to define adaptive constraints in $\varepsilon$-constraint method, which transforms the problem into a sequence of inequality-constrained subproblems, efficiently solved via constrained expected improvement acquisition. Our approach provides uniform Pareto coverage without hypervolume computation and naturally handles constraints and preferences. Experiments on synthetic and real-world benchmarks demonstrate superior coverage and competitive hypervolume performance against state-of-the-art baselines. Our code implementation can be found at https://github.com/YangYaohong1/STAGE-BO.
arXiv:2605.31367v1 Announce Type: new
Abstract: Token mixing layers play a key role in how language models can learn and generate long-range dependencies. Their efficiency relies on the necessary trade-off between decoding speed and the memory requirements, along with the cache size. Considering causal generation, this paper explores new trade-offs thanks to a unified framework which separates two crucial features: (i) the direct influence of inputs on outputs in one generation step; (ii) the recurrent propagation of information through past outputs. This framework encompasses major architectures such as attention and state-space models, but also generalizes the recurrence equations by allowing each state to depend on multiple past states rather than only the immediate predecessor. By introducing structure, we design new recurrence patterns that provably achieve the desired complexity, while providing theoretical insights on their expressivity -- trading runtime for expressivity in a principled way. Empirical validation is performed on synthetic tasks, along with language modeling. Together, these results provide a unified toolkit for the understanding and design of efficient and expressive token mixers across model families.
arXiv:2604.16922v3 Announce Type: replace
Abstract: Climate research is pivotal for mitigating global environmental crises, yet the accelerating volume of multi-scale datasets and the complexity of analytical tools have created significant bottlenecks, constraining scientific discovery to fragmented and labor-intensive workflows. While the emergence Large Language Models (LLMs) offers a transformative paradigm to scale scientific expertise, existing explorations remain largely confined to simple Question-Answering (Q&A) tasks. These approaches often oversimplify real-world challenges, neglecting the intricate physical constraints and the data-driven nature required in professional climate science.To bridge this gap, we introduce ClimAgent, a general-purpose autonomous framework designed to execute a wide spectrum of research tasks across diverse climate sub-fields. By integrating a unified tool-use environment with rigorous reasoning protocols, ClimAgent transcends simple retrieval to perform end-to-end modeling and analysis. To foster systematic evaluation, we propose ClimaBench, the first comprehensive benchmark for real-world climate discovery. It encompasses challenging problems spanning 5 distinct task categories derived from professional scenarios between 2000 and 2025. Experiments on ClimaBench demonstrate that ClimAgent significantly outperforms state-of-the-art baselines, achieving a 40.21% improvement over original LLM solutions in solution rigorousness and practicality. Our code are available at https://github.com/usail-hkust/ClimAgent.
arXiv:2604.17551v2 Announce Type: replace
Abstract: Standard approaches to goal-conditioned reinforcement learning (GCRL) that rely on temporal-difference learning can be unstable and sample-inefficient due to bootstrapping. While recent work has explored contrastive and supervised formulations to improve stability, we present a probabilistic alternative, called survival value learning (SVL), that reframes GCRL as a survival learning problem by modeling the time-to-goal from each state as a probability distribution. This structured distributional Monte Carlo perspective yields a closed-form identity that expresses the goal-conditioned value function as a discounted sum of survival probabilities, enabling value estimation via a hazard model trained via maximum likelihood on both event and right-censored trajectories. We introduce three practical value estimators, including finite-horizon truncation and two binned infinite-horizon approximations to capture long-horizon objectives. Experiments on offline GCRL benchmarks show that SVL combined with hierarchical actors matches or surpasses strong hierarchical TD and Monte Carlo baselines, excelling on complex, long-horizon tasks. Webpage and Code: https://simple-robotics.github.io/publications/survival-value-learning/
arXiv:2604.01985v2 Announce Type: replace
Abstract: General-purpose world models promise scalable policy evaluation, optimization, and planning, yet achieving the required level of robustness remains challenging. Unlike policy learning which primarily focuses on optimal actions, a world model needs to be reliable over a vast space of suboptimal actions, which are often underrepresented in action-labeled robot interactions. To address this challenge, we propose World Action Verifier (WAV), a framework that enables world models to identify their own prediction errors and self-improve. The key idea is to decompose action-conditioned state prediction into two independently verifiable factors: state plausibility and action reachability. We show that verifying these factors is significantly more tractable than direct forward prediction due to two underlying asymmetries: the broader availability of action-free data and the lower dimensionality of action-relevant features. Leveraging these asymmetries, we augment a world model with (i) a diverse subgoal generator obtained from video corpora and (ii) a sparse inverse model that infers actions from a subset of state features. By enforcing cycle consistency among proposed subgoals, inferred actions, and forward rollouts, WAV provides an effective verification mechanism in under-explored regimes, where existing methods often fail. Across nine tasks spanning MiniGrid, RoboMimic, and ManiSkill, our method achieves 2x higher sample efficiency while improving downstream policy performance by over 22%.
arXiv:2605.31360v1 Announce Type: new
Abstract: The Artificial Intelligence (AI) life cycle requires a thorough understanding of the underlying data dynamics for robust, safe and cost-effective AI development and use. Dataset shifts are defined as changes between train and test data distributions. Whether occurring over time (temporal) or across different sites (multi-source), they can severely degrade model performance and compromise data quality. This is particularly important in health AI, where the safety and fundamental rights of patients can be severely affected by uncontrolled shifts both at training and operational stages. While the theoretical foundations of covariate, prior, and concept shifts are well established, there is a lack of accessible and comprehensive software tools to perform their analysis. We introduce dashi, an open-source Python library designed for the exploration, quantification, and characterization of dataset shifts. dashi provides a dual approach: an unsupervised approach that leverages information geometry and non-parametric statistical manifolds to data variability characterization and analysis (e.g., Information Geometric Temporal plots and Multi-Source Variability metrics like Global Probabilistic Deviation and Source Probabilistic Outlyingness), and a supervised approach that quantifies and characterizes model performance degradation. Both unsupervised and supervised approaches work across user-defined temporal and domain/source batches. We demonstrate the utility of dashi on three simulated and real-world health AI case studies on gestational diabetes mellitus, COVID-19 and emergency medical dispatch. By providing interactive visual analytics and variability metrics, dashi supports trustworthiness of AI life cycle stages enabling robust and safe machine learning pipelines through the assessment of data coherence and AI performance.
Haptic Sorter: A Unified Planning Framework for Online Shape Estimation and Real-Time Pose Inference
arXiv:2605.31352v1 Announce Type: new
Abstract: Robotics manipulation usually assumes that the shape and pose of the object are known to the robot prior to motion planning. However, precise geometric information is not always available in practice, and pose inference suffers from sensor uncertainties and view occlusion. In this work, we propose a unified model-based geometric framework integrating robotic haptic perception, modeling, and manipulation planning. Our novelties involve: \textit{i)} Introducing Bayesian Optimization (BO) to guide the haptic exploration for object shape inference, where superellipses are used to approximate geometric boundary; \textit{ii)} Adaptive formulation of manipulation potential encoding object geometry for quasi-static robot-object interaction; \textit{iii)} Proposing an online Ordinary Differential Equation (ODE) for real-time pose inference based on model prediction and tactile feedback. We deploy our system on a 2D robotic sorting task, and vary object geometries to validate the robustness and generalizability of our framework in both simulation and a real-world multi-arm setup.
arXiv:2605.29994v2 Announce Type: replace
Abstract: 1D-CNNs play a crucial role for time-series analysis on tiny smart sensor systems, e.g. for biosignal analysis, predictive maintenance, or structural health monitoring. LUTbased precomputation has emerged as an interesting optimization technique to implement such neural networks on FPGAs. The core idea is to precompute all possible outputs of a neural network layer and store them directly in the lookup tables of the FPGAs. This enables highly resource-efficient networks with ultra-low latency but suffers from poor scalability. Previous work has explored using depthwise-separable convolutions to improve scalability. In this paper, we generalize this approach to consider additional forms of grouped convolutions. Based on this, we propose a novel type of convolutional block and an algorithm to guide the choice of hyper parameters for this block. We evaluate our approach on a medical time-series dataset for predicting atrial fibrillation using the MIT-BIH database (ECG recordings). The resulting hardware accelerators are small enough to be deployed on an AMD Spartan 7 S15. They achieve a F1-Score of up to 95% while only requiring 2,844 LUTs and no DSPs or BRAM.
arXiv:2509.12440v3 Announce Type: replace
Abstract: Deploying Large Language Models (LLMs) in medical applications requires fact-checking capabilities to ensure patient safety and regulatory compliance. We introduce MedFact, a challenging Chinese medical fact-checking benchmark with 2,116 expert-annotated instances from diverse real-world texts, spanning 13 specialties, 8 error types, 4 writing styles, and 5 difficulty levels. Construction uses a hybrid AI-human framework where iterative expert feedback refines AI-driven, multi-criteria filtering to ensure high quality and difficulty. We evaluate 20 leading LLMs on veracity classification and error localization, and results show models often determine if text contains errors but struggle to localize them precisely, with top performers falling short of human performance. Our analysis reveals the "over-criticism" phenomenon, a tendency for models to misidentify correct information as erroneous, which can be exacerbated by advanced reasoning techniques such as multi-agent collaboration and inference-time scaling. MedFact highlights the challenges of deploying medical LLMs and provides resources to develop factually reliable medical AI systems.
arXiv:2605.31339v1 Announce Type: new
Abstract: Periodic network-wide dissemination of node location data is crucial for shared situational awareness and collaborative mapping in mobile ad hoc and mesh networks for public safety, disaster relief, and military. A key challenge is to provide maximally accurate location information with minimal energy expenditure on part of the nodes. We present SQEEZ: a mechanism for reducing the Position Location Information (PLI) load that combines two orthogonal techniques: (1) adaptive suppression of location updates; and (2) temporal and inline compression of update packets. We describe the SQEEZ suppression and compression algorithms, analyze the tradeoff between location error and energy consumption, and introduce a new metric called \textit{Error-Penalized-Energy (EPE)} that normalizes the energy metric using the error incurred. Our simulation results show that, in the range of parameters studied, SQEEZ improves the EPE-efficiency and scalability in a 30-node random waypoint scenario by up to 4.4x and 2.3x respectively; and increases the EPE-efficiency by 7.5x in a 9-node real-world network trace. Compression provides larger improvements than suppression at high mobilities and vice-versa at low mobilities.
arXiv:2601.01456v2 Announce Type: replace
Abstract: In this paper, we revisit multimodal few-shot 3D point cloud semantic segmentation (FS-PCS), identifying a conflict in "Fuse-then-Refine" paradigms: the "Plasticity-Stability Dilemma." In addition, CLIP's inter-class confusion can result in semantic blindness. To address these issues, we present the Decoupled-experts Arbitration Few-Shot SegNet (DA-FSS), a model that effectively distinguishes between semantic and geometric paths and mutually regularizes their gradients to achieve better generalization. DA-FSS employs the same backbone and pre-trained text encoder as MM-FSS to generate text embeddings, which can increase free modalities' utilization rate and better leverage each modality's information space. To achieve this, we propose a Parallel Expert Refinement module to generate each modal correlation. We also propose a Stacked Arbitration Module (SAM) to perform convolutional fusion and arbitrate correlations for each modality pathway. The Parallel Experts decouple two paths: a Geometric Expert maintains plasticity, and a Semantic Expert ensures stability. They are coordinated via a Decoupled Alignment Module (DAM) that transfers knowledge without propagating confusion. Experiments on popular datasets (S3DIS, ScanNet) demonstrate the superiority of DA-FSS over MM-FSS. Meanwhile, geometric boundaries, completeness, and texture differentiation are all superior to the baseline. The code is available at: https://github.com/MoWenQAQ/DA-FSS/.
arXiv:2509.20941v2 Announce Type: replace
Abstract: As surgical AI transitions from pixel-level detection to complex reasoning, Scene Graphs (SGs) offer the structured, relational representations necessary to decode dynamic surgical environments. This PRISMA-ScR-guided scoping review systematically maps the evolving landscape of SG research in surgery, analyzing 52 primary studies to chart applications and methodological shifts. Our analysis reveals rapid growth, yet uncovers a critical 'data divide': internal-view research (e.g., triplet recognition from endoscopic video) accounts for 81% of studies and almost exclusively uses real-world 2D video, while external-view operating room modeling relies heavily on simulated data. Methodologically, we identify a decisive shift from foundational graph neural networks to specialized foundation models and generative AI, which together now account for approximately 50% of research in 2025. Crucially, our synthesis suggests that Scene Graphs are evolving from simple descriptors into essential 'neuro-symbolic guardrails', providing the structured, verifiable intermediate representation needed to prevent hallucinations in increasingly autonomous Surgical Foundation Models. Despite this promise, a major translational gap remains: none of the reviewed studies have proceeded to prospective clinical validation. We conclude that bridging this gap requires moving beyond standard computer vision metrics; we therefore propose the 'Validation Trinity' -- prioritizing Semantic Query Success, Latency-Aware Accuracy, and Safety-Critical Recall -- as the necessary evaluation framework to bring graph-based surgical AI into clinical practice.
arXiv:2510.00419v2 Announce Type: replace
Abstract: Zeroth-order optimizers have recently emerged as an attractive approach for fine-tuning large language models (LLMs), as they avoid backpropagation and can substantially reduce memory overhead relative to standard first-order training. However, existing zeroth-order methods rely on hand-crafted, static sampling strategies that are not adaptable to model-specific structures. To address this, we propose ZO-Finetuner, a learning-based zeroth-order optimizer for LLMs that automatically learns efficient perturbation strategies through a compact and memory-efficient design. Motivated by the fact that a small set of base LLMs is repeatedly fine-tuned across tasks, ZO-Finetuner supports one-time per-model training and reuse across downstream tasks with minimal overhead. Therefore, learning the optimizer once for a given LLM and reusing it across diverse downstream tasks is both feasible and highly desirable. Accordingly, ZO-Finetuner is designed to scale learning to learn (L2L) to the foundation-model era by supporting one-time per-model training with minimal overhead. Experiments on 4 LLMs and 7 datasets show that ZO-Finetuner outperforms prior zeroth-order baselines in 82.1\% of task-model combinations, thereby demonstrating strong performance and scalability for efficient LLM fine-tuning. The code can be found in https://github.com/ASTRAL-Group/ZO_Fine_tuner.
arXiv:2605.30684v1 Announce Type: new
Abstract: Emergent constraints in climate science are empirical relations that link the response to a forcing of a physical observable to the properties of other observables, with the aim of reducing climate change projection uncertainties. Here we use recent results in linear response theory to develop a mathematical framework for dynamic emergent constraints, a class of emergent constraints linking the response of different observables to the same forcing. We show how traditional dynamic emergent constraints are a special case of more general relations, that we call integral dynamic emergent constraints. These relations allow to compute the response of a predictand as the convolution of the response of a predictor and the proxy Green's function of the predictand-predictor pair. The conditions for the existence of integral emergent constraints are related to the causality of the proxy Green's function and the time scales at which the system is observed. We apply this framework to global warming simulations with the MPI-ESM climate model, to study dynamic emergent constraints between different observables. These results allow to put the theory of dynamic emergent constraints on firm mathematical ground, and suggest a protocol to identify necessary conditions for the existence of such relations in climate data.