arXiv:2607.12085v1 Announce Type: new
Abstract: Evaluating retail conversational agents requires methods beyond lexical-overlap metrics to assess intent alignment, factuality, helpfulness, clarity, tone, and overall response quality. Although LLM-as-a-judge methods provide scalable alternatives to human evaluation, production deployment introduces challenges in governance, reproducibility, cost, schema consistency, traceability, and reliability. We present GenAI Evaluation, a governed, configuration-driven pipeline for large-scale evaluation of retail conversational systems. It processes production chatbot logs through normalization, sharding, asynchronous execution, and schema-constrained LLM scoring. The framework evaluates helpfulness, truthfulness, clarity, tone alignment, and translation-specific dimensions. Selective re-evaluation processes only incomplete, malformed, or schema-invalid records, while schema locking, versioned configurations, validation logs, and record-level provenance support auditability. The framework processes approximately 50,000 records daily and has evaluated more than two million interactions. Validation used 12,980 stratified-random human-labeled records from four trained annotators. Classification covered 14 intents, 156 sub-intents, 18 major domains, and 129 sub-domains. The pipeline achieved a macro F1 score of 0.93 and 89% human-acceptability accuracy for translation.
Science Journals
arXiv:2607.11988v1 Announce Type: cross
Abstract: The classical Gibbs-Duhem relation applies to quasi-static processes and neglects kinetic effects, leaving a fundamental gap between Gibbs thermodynamics and Newtonian mechanics. Here, we derive a generalized Gibbs-Duhem framework that incorporates kinetic contributions, thereby establishing a unified connection between classical thermodynamics and Newtonian mechanics. Based on this framework, we propose an alternative evolution equation governing density dynamics at fluid-fluid interfaces. In appropriate limiting cases, the resulting density evolution equation naturally recovers the definition of the speed of sound, Bernoulli's law, and the van der Waals equation of state (EOS).
arXiv:2607.12586v1 Announce Type: cross
Abstract: Medical image segmentation is a crucial task in the field of clinical analysis and applications. Though deep learning techniques recently play a crucial role in several scenarios, the training at the individual pixel level leads to a lack of geometric prior information. Scholars proposed to integrate the Chan-Vese model into the loss function for training which can take into account the region and length of the region inside and outside the segmentation process and then improve the performance in medical image segmentation. However, these methods still lack an effective characterization of the segmented region. To overcome this problem, we introduce the mean curvature as a geometric natural constraint and propose a Deep Active Contour and Mean Curvature (DACMC) loss function where the convolution kernel is used to approximate the mean curvature to save computational cost. We have validated the performance of our method on the liver and spleen dataset. Our proposed method demonstrates new state-of-the-art performance on several segmentation datasets.
arXiv:2607.12676v1 Announce Type: cross
Abstract: Inelastic light scattering is a central tool for sample characterization and label-free imaging across the physical and life sciences. Recent work has suggested that scattered light can also exhibit nonclassical correlations. Here, we develop a microscopic theory of Raman scattering that connects naturally to established descriptions of entangled-photon generation in three- and four-wave mixing, while capturing essential differences arising from the resonant, dissipative character of Raman scattering. Using a cumulant-expansion approach, we analyze the entanglement structure of the scattered sidebands and identify signatures that distinguish Raman-mediated correlations from those generated in conventional off-resonant nonlinear wave mixing. In particular, we show that Raman scattering induces chiral couplings between Stokes and anti-Stokes sidebands, leading to nonreciprocal amplification of coherent seed fields. These results establish a theoretical framework for Raman-based quantum photonic protocols and suggest routes toward quantum-enhanced Raman spectroscopy and imaging.
arXiv:2607.12089v1 Announce Type: new
Abstract: Large language models (LLMs) frequently generate code with security vulnerabilities, yet these weaknesses are rarely isolated: they often span multiple concern areas simultaneously, reflecting the cross-cutting nature of security in software. We present a framework that combines security-oriented Metamorphic Relations (MRs) with Association Rule (AR) mining to detect vulnerabilities in LLM-generated code, uncover their co-violation structure, and trace that structure back to prompt-level risk factors. We define nine MRs covering major CWE categories, including SQL injection, XSS, command injection, path traversal, hard-coded credentials, weak cryptography, and memory-safety errors, and apply them using an LLM-based judge to 3,700 code snippets generated by five open models from the LLMSecEval benchmark. The results show that 68.8% of snippets violate at least one MR, with hard-coded credentials (79.1%) and command injection (74.4%) among the most prevalent applicable failures. AR mining reveals strong cross-cutting co-violation patterns, notably that XSS and weak cryptography co-violations predict hard-coded credentials with 82.5% confidence (lift = 3.23), along with tightly coupled clusters linking authentication, credential handling, and cryptographic weakness, as well as input-handling and memory-safety failures. We then perform prompt-level risk analysis and find that database- and authentication-related prompts are strong predictors of broad cross-cutting insecurity, while 65.5% of prompts yield consistent violation outcomes across all five models. These findings show that insecure code generation is not merely a collection of independent defects, but a structured and prompt-conditioned phenomenon, motivating cluster-aware verification and prompt-level intervention for safer LLM-assisted programming.
arXiv:2607.12122v1 Announce Type: new
Abstract: We present an agentic approach to autonomous neural operator discovery based on an AI scientific community, which consists of a swarm of virtual laboratories that interact under a citation-based economy of influence. Highly-cited labs found new labs that follow their research direction and replace non-performing labs. Each virtual lab contains three agents: an LLM planner that proposes an architecture, a numerical worker that trains and measures it, and an LLM reviewer that participates in cross-lab peer review. All labs share a common vocabulary consisting of DeepONet (branch-trunk), Fourier, Transformer (attention), wavelet, and residual convolutional neural operator building blocks. We evaluate the neural operator AI scientific community on five problems, namely piecewise regression, the linear advection and Burgers 1D PDEs, and the Navier-Stokes and Darcy flow 2D PDEs, while repeating the simulation three times for each problem. The results show that the neural operator AI scientific community is capable of discovering high-accuracy, low-parameter-count neural operator architectures. All 9,623 LLM calls are logged and audited, which reveals that the virtual lab LLM planners choose to hybridize in 99.8% of their logged decisions, consistently returning multi-family hybrids. Moreover, we conducted an ablation study by replacing the LLM agents in each lab by rule-based alternatives, which caused the scientific community to collapse to non-hybridized single-family stacks in several cases, showing that LLM agency is needed to preserve diversity. The results suggest a no-free-lunch theorem for neural operators: there is no universal winner. The code, configurations, and the complete LLM transcripts are released at https://github.com/luislootx/AI-SC.
arXiv:2405.19466v5 Announce Type: replace
Abstract: We pose uncertainty quantification and exploration in online decision-making as a problem of training and generation from an autoregressive sequence model, an area experiencing rapid innovation. Our approach rests on viewing uncertainty as arising from missing future outcomes that could be revealed through action choices, rather than from unobservable latent parameters of the environment. This reformulation aligns naturally with modern machine learning capabilities: we can i) train generative models through next-outcome prediction rather than fit explicit priors, ii) assess uncertainty through autoregressive generation rather than sampling latent parameters from posteriors, and iii) adapt to new information by extending the sequence model's context rather than explicit posterior updating. Our main theoretical result establishes a reduction from online decision-making to offline next-outcome prediction: Bayesian regret is controlled directly by the sequence model's offline prediction loss, without requiring an explicit latent-variable posterior. Experiments, including a semi-synthetic news recommendation task, show that autoregressive generation produces calibrated epistemic uncertainty and enables effective exploration by using article text as prior information to focus exploration on resolving remaining uncertainties.
arXiv:2607.12169v1 Announce Type: new
Abstract: Intercomprehension refers to partial intelligibility of an unfamiliar language (L2) by a speaker of a related language (L1). How is this zero-shot cross-language comprehension possible? In this work, we extend past work on algorithmic models of noisy-channel inference to model intercomprehension in a Bayesian framework. The model uses an LM in L1 only for scoring latent hypotheses about the translations of observed L2 utterances, and a general-purpose noise model to infer a mapping between L2 and L1 words based on either form-based similarity or symbolic rules. We then conduct a human behavioral experiment, eliciting inferences for utterances in Dutch, Italian, and Ukrainian from speakers of English, Spanish, and Russian, respectively. Our full model shows a closer alignment to the distribution of human intercomprehension performance than ablations, and also compares favorably to zero-shot prompting of much larger models. These results provide a cognitively plausible computational model of intercomprehension, and highlight the flexible inferences made by comprehenders under wide uncertainty in real-world cross-language scenarios. We share our code publicly.
arXiv:2607.12186v1 Announce Type: new
Abstract: In-loop filters have been comprehensively explored during the development of video coding standards due to their remarkable noise-reduction capability. In the early stage of video coding, in-loop filters, such as Deblocking Filter, Sample Adaptive Offset, and Adaptive Loop Filter, were performed separately for each component. Recently, cross-component filters were studied to improve the chroma fidelity by exploiting correlations between the luma and chroma channels. This paper summarizes the cross-component filters used in the state-of-the-art video coding standard. Specifically, it includes the Cross-Component Adaptive Loop Filter and Cross-Component Sample Adaptive Offset. Cross-component filters aim to reduce compression artifacts based on the correlation between different components and provide more accurate pixel reconstruction values. In this paper, we introduce the origin, development, and status of cross-component filters in the current video coding standards. Finally, we had some discussions on the further evolutions of cross-component filters.
arXiv:2509.15120v2 Announce Type: replace
Abstract: In high-stakes scenarios, such as medical imaging applications, it is critical to equip the predictions of a regression model with reliable confidence intervals. Recently, Conformal Prediction (CP) has emerged as a powerful statistical framework that, based on a labeled calibration set, generates intervals that include the true labels with a pre-specified probability. In this paper, we address the problem of applying CP for regression models when the calibration set contains noisy labels. We begin by establishing a mathematically grounded procedure for estimating the noise-free CP threshold. Then, we turn it into a practical algorithm that overcomes the challenges arising from the continuous nature of the regression problem. We evaluate the proposed method on two medical imaging regression datasets with Gaussian label noise. Our method significantly outperforms the existing alternative, achieving performance close to the clean-label setting.
arXiv:2509.24372v3 Announce Type: replace
Abstract: Fine-tuning large language models (LLMs) for downstream tasks is an essential stage of modern AI deployment. Reinforcement learning (RL) has emerged as the dominant fine-tuning paradigm, underpinning many state-of-the-art LLMs. In contrast, evolution strategies (ES) has largely been overlooked due to the widespread belief that it does not scale to modern model sizes. This paper overturns this assumption by demonstrating the first successful application of ES to full-parameter fine-tuning of LLMs at the billion-parameter scale, without dimensionality reduction. ES can indeed search over extremely high-dimensional parameter spaces and outperform established RL implementations across multiple axes, including improved tolerance to long-horizon and delayed rewards, robustness across diverse base LLMs, reduced susceptibility to reward hacking, and improved training stability. These findings suggest that ES is not merely a viable alternative to RL, but a fundamentally different and powerful backpropagation-free post-training paradigm that opens a new direction for LLM fine-tuning beyond current RL-based approaches.
arXiv:2509.24863v2 Announce Type: replace
Abstract: Deep neural networks surpass humans on many vision benchmarks, yet remain far less robust to distribution shifts such as illumination and weather changes. Existing approaches address this challenge by additional training data, extensive augmentation, architectural modifications, or test-time adaptation. In this work, we explore a complementary direction: inspired by the human retina, we propose a fixed, model-agnostic preprocessing module that extracts signals that are more stable with respect to variations of illumination. Our method combines color remapping with local contrast extraction, producing sparse representations that emphasize structural features. We study its impact on semantic segmentation by training on Cityscapes and evaluating generalization under adverse conditions on Dark Zurich and ACDC. Our results show that the biologically inspired preprocessing preserves in-distribution performance while consistently improving robustness in challenging lighting scenarios, such as nighttime, where annotated training data are scarce. Moreover, the segmentation accuracy remains stable even when the contrast-based representation is sparsified by up to 70%. These gains suggest that rethinking the input representation itself can improve robustness while also opening opportunities for lower-latency, transmission-aware imaging sensors when sparsity can be exploited close to acquisition.
arXiv:2607.12271v1 Announce Type: new
Abstract: In this paper, a parametric physics-informed neural network for solving the heterogeneous soil thermal problem with borehole heat exchangers (BHEs) as singular sources is developed. There are three novel features in the present framework; namely, (i) the singularity is naturally removed by using analytical line source models; (ii) using the explicit formulation for gradient thermal conductivity enables physics-informed learning of the parametrization featuring the conductivity; (iii) the learned correction is utilized as an efficient universal corrector via superposition principles. We first introduce the decomposition of the temperature change and transform the approximation of the entire heterogeneous response to the correction compensating the difference between the practical solution and idealized homogeneous approximation. In such a way, the delta function singularity is excluded and the bulk heat transfer is captured for the sake of facilitating the effective training of the neural network. The original problem is then reformulated as a governing correction diffusion or advection-diffusion equation subject to a homogeneous initial condition. The linearly varying thermal conductivity is used to model the soil heterogeneity. We propose a physics-informed neural network to approximate a universal corrector with respect to a single borehole with unit heat extraction rate. As a result, the network is trained by minimizing the physics-informed and data-anchored loss function that is evaluated for sampled conductivity parameters on adaptively selected training points. In addition, we include the location indicator function regarding the source as a feature input of network and find that it helps the network to process the local information. We perform numerical tests to exhibit the effectiveness of the proposed method based on three different analytical models.
arXiv:2607.12272v1 Announce Type: new
Abstract: When the guiding-center description fails and the full gyromotion must be resolved for energetic particles in complex configurations like stellarators, charged-particle integrators must be formulated directly in the curvilinear flux coordinates. The Boris algorithm, which adopts a staggered scheme in Cartesian coordinates, is phase-space-volume-preserving and second-order accurate; but a direct port to flux coordinates degrades the position update to first order, because the evolving basis vectors of the curvilinear frame make the starting-point metric deviate from the ideal midpoint metric. We construct a collocated, midpoint-predicted Boris algorithm in flux coordinates, restoring second-order accuracy at the cost of one additional field evaluation per step. In reactor-scale stellarator magnetic fields, the scheme recovers second-order convergence in every coordinate component, retains near-machine-precision energy conservation and a bounded magnetic moment, and demonstrates greater orbit robustness than Staggered Boris and RK4 at coarse time steps.
arXiv:2603.08465v3 Announce Type: replace
Abstract: While Physics-Informed Neural Networks (PINNs) offer a mesh-free approach to solving fluid-flow PDEs, standard point-wise residual minimization suffers from convergence pathologies in topologically complex domains like Triply Periodic Minimal Surfaces (TPMS). The locality bias of point-wise constraints fails to propagate global information through tortuous channels, causing unstable gradients and conservation violations. To address this, we propose the Multi-scale Weak-form PINN (MUSA-PINN), which reformulates Navier-Stokes equation constraints as integral conservation laws over hierarchical spherical control volumes. We enforce continuity and momentum conservation via flux-balance residuals on control surfaces. Our method utilizes a three-scale subdomain strategy-comprising large volumes for long-range coupling, skeleton-aware meso-scale volumes aligned with transport pathways, and small volumes for local refinement-alongside a two-stage training schedule prioritizing continuity. Experiments on steady incompressible flow in TPMS geometries show MUSA-PINN outperforms state-of-the-art baselines, reducing relative errors by up to 93% and preserving mass conservation.
arXiv:2603.14217v3 Announce Type: replace
Abstract: In cognitive science and linguistic theory, dialogue is not seen as a chain of independent utterances but rather as a joint activity sustained by coherence, consistency, and shared understanding. However, many systems for open-domain and personalized dialogue use surface-level similarity metrics (e.g., BLEU, ROUGE, F1) as one of their main reporting measures, which fail to capture these deeper aspects of conversational quality. We re-examine a notable retrieval-augmented framework for personalized dialogue, LAPDOG, as a case study for evaluation methodology. Using both human and LLM-based judges, we identify limitations in current evaluation practices, including corrupted dialogue histories, contradictions between retrieved stories and persona, and incoherent response generation. Our results show that human and LLM judgments align closely but diverge from lexical similarity metrics, underscoring the need for cognitively grounded evaluation methods. Broadly, this work charts a path toward more reliable assessment frameworks for retrieval-augmented dialogue systems that better reflect the principles of natural human communication.
arXiv:2605.04517v2 Announce Type: replace
Abstract: Gausslets are one of the few basis constructions for electronic structure that combine locality, orthonormality, variable resolution, and an accurate diagonal approximation for the electron-electron interaction, but the original construction is tied to one dimension. Radial gausslets extended this idea to atoms while leaving the angular degrees of freedom in spherical harmonics, so the atomic interaction remained only partially diagonal in the combined basis. Here we introduce generalized gausslets on the sphere and combine them shell by shell with radial gausslets to form an atom-centered basis in which the electron-electron interaction takes a two-index integral-diagonal form. The angular basis starts from localized spherical Gaussians and uses injection to make a low-$\ell$ spherical-harmonic subspace exact. Tests of the kinetic spectrum, low-$\ell$ Coulomb matrix elements, spherium, first-row Hartree--Fock calculations, and He exact diagonalization show systematic convergence with increasing angular resolution. We also develop DMRG methods for this basis, including compact MPOs, correlated small-space starting states, Givens-rotation transfers between nearby angular sizes, and embedded sampled variance extrapolation (ESVE). We show that this combination of ingredients can be used to solve the Be atom, with extrapolations in the number of angular functions but with fixed radial resolution, to within about 0.1 mH of the complete basis set limit exact energy. This shows that DMRG calculations of first row atoms which include both static and accurate dynamic correlation on the same footing are feasible.
arXiv:2605.31582v2 Announce Type: replace
Abstract: Strongly correlated electrons can be treated with a configuration interaction of Slater determinants grouped by number of unpaired electrons with exponential cost. The first two papers in this series demonstrated that single reference methods built from Richardson-Gaudin states gave results of similar quality at polynomial cost. In this contribution, the states are simplified substantially yielding the perfect-pairing state as a reference along with its low-lying excitations. The states are much simpler, the computational cost is substantially reduced, and there is no sacrifice in numerical accuracy. Second-order Epstein-Nesbet perturbative corrections for the valence electrons are similar in quality to the complete active space self-consistent field.
arXiv:2607.12130v1 Announce Type: new
Abstract: Spacecraft attitude control is traditionally achieved using momentum exchange devices or propellant-consuming thrusters. Meanwhile, a growing number of missions require robotic manipulators, which are typically treated as disturbance sources to be rejected rather than as actuators for spacecraft reorientation. This work investigates the use of manipulator motions for propellant-free attitude control by formulating a trajectory optimization problem with critical joint and collision avoidance constraints. Using an interior point solver for the resulting nonlinear program, complex slew and detumble trajectories are demonstrated for a range of spacecraft-manipulator systems with varying kinematic complexity and mass properties. The achievable control authority is compared directly with that of reaction wheel arrays via momentum and torque envelopes, demonstrating the potential for manipulators to serve as redundant or even primary attitude control systems. This work provides a framework for using manipulators as multipurpose attitude control actuators, with particularly promising applications in in-space assembly and manufacturing when grasping payloads with high relative mass fractions.
arXiv:2607.12311v1 Announce Type: new
Abstract: Self-accelerating optical beams follow curved trajectories rather than propagating rectilinearly, which raises the prospect for avoiding line-of-sight (LoS) obstacles blocking the beam path. However, if a conventional laser beam is intercepted by an obstacle en route to an intended LoS target, then replacing this beam with a bending beam is not adequate: the obstacle is avoided but the target cannot be concomitantly reached. Rather, a laser beam that \textit{locally} bends around the obstacle -- before continuing along its rectilinear path -- is needed. Here we show that engineering the spatiotemporal spectrum of a pulsed beam yields a space-time wave packet whose propagation dynamics can be tuned at will to locally bend around one or multiple obstacles, thereby avoiding them, to selectively reach a designated target. We carry out our experiments over distances extending to 130~m from the source, carried out in an open-air environment. In one scenario, the beam is incident on a target placed between two LoS obstacles, one preceding it and one following it -- both of which are avoided. In a second scenario, the locally bending beam avoids one and then two LoS obstacles preceding the intended target. These results may contribute to applications requiring selective incidence on targets in remote sensing, stand-off detection, directed energy, for optical and radio-frequency communications in presence of LoS obstacles, and for selectively delivered radiation therapies.
arXiv:2607.12193v1 Announce Type: new
Abstract: While generative AI has unlocked new opportunities for 3D content creation, current workflows often rely on multiple regenerations, which provides limited control and unpredictable outcomes. We present Compos3D, a system that introduces a compositional workflow for generative 3D modeling through remixing. Instead of repeatedly regenerating models, users generate multiple candidates from text or image prompts, select parts of interest via 2D image regions or 3D mesh segments, and assemble them into a coherent design. The system synthesizes these compositions into a refined 3D model, preserving high-level intent while resolving low-level geometry. To evaluate this approach, we conducted a controlled user study comparing remixing and regeneration workflows across both 2D and 3D modalities. Results show that the remixing workflow provides participants with greater creative control, stronger alignment with their intent, and higher satisfaction. We conclude with design recommendations for future AI-assisted 3D modeling workflows.
arXiv:2607.12266v1 Announce Type: new
Abstract: Mixed-precision quantization must decide which parts of a model to keep at higher precision. A common premise, shared by sensitivity-based methods such as HAWQ and CoopQ, is that the loss from quantizing a set of layers can be reconstructed from per-layer or pairwise sensitivities measured in isolation. We test this premise at the 4-bit weight-and-activation precisions now being deployed, treating the change in loss $f(S)$ from quantizing a layer set $S$ as a set function on the Boolean cube and analyzing it through two classical changes of basis. This analysis yields two findings. First, across configurations drawn from the deployment distribution, 85--93\% of the variance of $f$ is explained by per-layer effects alone. Second, a monotone transform of a sum of per-layer terms reproduces $f$'s ranking of configurations, misordering at most 2\% of pairs.
We propose the coverage model $f(S)=c\bigl(1-\prod_{i\in S}(1-a_i)\bigr)$, which reproduces the measured variance profile of $f$ to within a few percent from its $L$ fitted break-rates. This structure supports two predictors of a configuration's loss, each with $L+1$ parameters. The additive model is the optimal first-order predictor. By Parseval's identity its mean-squared error equals the variance of $f$ left unexplained by per-layer effects, which we measure on full lattices, estimate out of sample at full-network scale, and report with every result as a certificate of how well any additive model can do. The coverage model itself is the second predictor.
As allocators at matched memory, they attain the lowest KL divergence among the compared allocators on models from 30B to 355B parameters. Below four bits, the resulting allocations continue to solve code and reasoning tasks at budgets where allocations from gradient sensitivities no longer produce terminating generations.
arXiv:2607.12343v1 Announce Type: new
Abstract: This paper studies learning-based MPC for constrained stabilization of discrete-time linear systems with unknown system parameters and additive bounded disturbances. We develop a tractable homothetic-tube MPC scheme in which a high-probability parameter confidence set is generated from non-asymptotic regularized least-squares estimation, rather than assumed a priori. The resulting uncertainty set is embedded into robust tube propagation and constraint tightening, yielding a convex formulation with linear and second-order-cone constraints. We prove high-probability recursive feasibility, robust constraint satisfaction, and input-to-state stability, together with explicit non-asymptotic state bounds. A numerical example illustrates the effectiveness and theoretical guarantees.
arXiv:2602.12783v3 Announce Type: replace
Abstract: Spoken query retrieval is an important interaction mode in modern information retrieval. However, existing evaluation datasets are often limited to simple queries under constrained noise conditions, making them inadequate for assessing the robustness of spoken query retrieval systems under complex acoustic perturbations. To address this limitation, we present SQuTR, a robustness benchmark for spoken query retrieval that includes a large-scale dataset and a unified evaluation protocol. SQuTR aggregates 37,317 unique queries from six commonly used English and Chinese text retrieval datasets, spanning multiple domains and diverse query types. We synthesize speech using voice profiles from 200 real speakers and mix 17 categories of real-world environmental noise under controlled SNR levels, enabling reproducible robustness evaluation from quiet to highly noisy conditions. Under the unified protocol, we conduct large-scale evaluations on representative cascaded and end-to-end retrieval systems. Experimental results show that retrieval performance decreases as noise increases, with substantially different drops across systems. Even large-scale retrieval models struggle under extreme noise, indicating that robustness remains a critical bottleneck. Overall, SQuTR provides a reproducible testbed for benchmarking and diagnostic analysis, and facilitates future research on robustness in spoken query to text retrieval.
arXiv:2605.09977v3 Announce Type: replace
Abstract: Spatio-temporal fetal brain atlases are important for characterizing normative neurodevelopment and identifying congenital anomalies. However, existing atlas construction pipelines necessitate days for slice-to-volume reconstruction (SVR) to generate high-resolution 3D brain volumes and several additional days for iterative volume registration, thereby rendering atlas construction from large-scale cohorts prohibitively impractical. We address these limitations with INFANiTE, an Implicit Neural Representation (INR) framework for high-resolution Fetal brain spatio-temporal Atlas learNing from clinical Thick-slicE MRI scans, bypassing both the costly SVR and the iterative non-rigid registration steps entirely, thereby substantially accelerating atlas construction. Extensive experiments demonstrate that INFANiTE outperforms existing baselines in subject consistency, reference fidelity, intrinsic quality and biological plausibility, even under challenging sparse-data settings. Additionally, INFANiTE reduces the end-to-end processing time (i.e., from raw scans to the final atlas) from days to hours compared to the traditional 3D volume-based pipeline (e.g., SyGN), facilitating large-scale population-level fetal brain analysis. Code: https://github.com/hu2274898/INFANiTE