// intelligence · research
AI Research Papers
Preprints and publications from arXiv, Berkeley AI Research, MIT, Google Research, and leading academic institutions worldwide.
500 papers · showing 20
From Log-Odds to Shapley Values: An Explanatory Geometry for the Weighted Naive Bayes Classifier
arXiv:2610.10642v1 Announce Type: new Abstract: This paper studies the construction of an explanatory space for a weighted naive Bayes classifier from the supervised representation induced by the model. We start from the classical supervised distance based on conditional log-likelihoods and introduce a discriminative reformulation based on log-odds, which is more directly related to the classification decision. We then show that this representation induces a distance that exactly coincides with the $\ell_1$ distance between vectors of analytical Shapley values, thereby providing a formal explanatory interpretation of the geometry induced by the model. Finally, we empirically compare several supervised distances derived from these representations using a $k$-nearest neighbors classifier. This work highlights a close link between supervised distance, local explanation, and predictive behavior, from a primarily methodological perspective.
Polynomial $\mathcal{Z}$-contraction: A unified approach with Application
arXiv:2610.10546v1 Announce Type: new Abstract: In this paper, we introduce polynomial $\mathcal{Z}$-contraction by unifying the ideas of polynomial contraction and $\mathcal{Z}$-contraction and also provide demonstrative examples to show the effectiveness and novelty. Chiefly, we prove two fixed point results for polynomial $\mathcal{Z}$-contraction employing simulation functions. Lastly, we utilize our main result to discuss the existence and uniqueness of solution for boundary value problem with a concrete example in $L^p(0,\pi)$ under suitable assumptions.
Interpretable Memory Models for Spaced Repetition
arXiv:2610.10548v1 Announce Type: new Abstract: Spaced repetition software schedules reviews with a memory model fit to review logs. Accuracy on a test set is not sufficient evidence of quality since available data are produced by existing schedulers, and new solutions must extrapolate beyond them. A model also needs a simple mechanistic interpretation. We present SBD, a model that is more interpretable and 80% smaller than the current state of the art at nearly the same accuracy.
Wells exact sequences for $n$-Lie algebras
arXiv:2610.10561v1 Announce Type: new Abstract: We study derivations and automorphisms of arbitrary non-abelian extensions of $n$-Lie algebras through their full crossed-product data. Their relative center is a canonical coefficient module, and a third cohomology class obstructs completion of data satisfying the mixed Filippov identities. When this class vanishes, a relative second cohomology space classifies the remaining extension data. We prove a non-abelian good criterion, construct the Lie algebra and group of compatible pairs, and obtain derivation and automorphism Wells sequences with explicit lifting equations. The derivation obstruction is a Lie algebra $1$-cocycle; the additive automorphism obstruction is a crossed homomorphism. We distinguish the ordinary center from the relative center, and outer derivations of the kernel from the mixed action required. Sparse crossed products are treated, including the representation class of their second mixed operation and the additional equations governing changes of sparse sections.
Inflationary Dynamics and CMB Constraints in $f(T,L_m)$ gravity
arXiv:2610.10565v1 Announce Type: new Abstract: We investigate slow-roll inflation in a modified teleparallel framework with a non-minimal coupling between torsion and the matter sector. Focusing on Starobinsky-type plateau inflation, we derive the modified background dynamics and the corresponding slow-roll observables, and examine how the gravitational couplings affect the scalar spectral index, tensor-to-scalar ratio, and running of the spectral index. The model continuously recovers the standard Starobinsky scenario in the appropriate weak-coupling limit, while the matter--geometry coupling produces a non-monotonic modification of the inflationary predictions. We find that weak couplings leave the standard predictions essentially unchanged, whereas stronger couplings can significantly shift the scalar spectral index and eventually become incompatible with current cosmic microwave background constraints. The quadratic torsion correction has a comparatively mild impact within the controlled low-energy regime, primarily affecting the tensor amplitude. We confront the resulting predictions with the Planck and BICEP/Keck constraints and find that the model remains viable over a broad range of the inflationary parameter space. Our results demonstrate that the matter--geometry coupling can leave a potentially testable imprint on primordial observables while retaining the successful plateau-inflation phenomenology of the Starobinsky scenario.
Personalization Matters: Long-Horizon Conversation Agent with User-Centric Information in Online Shopping Interactions
arXiv:2610.11375v1 Announce Type: new Abstract: Personalized conversational shopping requires maintaining preference consistency over multi-turn interactions, where users reveal constraints gradually. Existing approaches often rely on static profiles and do not explicitly control long-horizon interaction behavior. We propose a multi-agent, multimodal Retrieval-Augmented Generation (RAG) framework that decomposes dialogue state tracking, recommendation retrieval, preference-aware reasoning, and response generation, while integrating product metadata, product reviews, image-derived descriptions, and user historical reviews. To evaluate interaction-level quality, we adopt a trajectory-level protocol with four dimensions: Global Preference Consistency, Cumulative Information Synthesis, Interaction Trajectory, and Tone Consistency. On an Amazon Reviews 2023 benchmark, retrieval-enabled variants outperform a no-RAG baseline on automatic trajectory metrics (average 4.82 vs. 3.74). In a small real-user study ($n{=}5$), the Full variant achieves the highest mean overall rating (4.60 vs. 2.20 for Baseline), providing exploratory evidence that role decomposition plus user-centric retrieval improves perceived personalization.\footnote{Code and dataset are available at: https://github.com/RenaGao/Multimodel_RAG_Indexing
TacHair: Tactile Contact-Distribution Guided Online Correction for Robotic Hair Stroking and Perception
arXiv:2610.10637v1 Announce Type: new Abstract: Hair stroking is common in daily grooming and personal care, and is also widely used in hair-product evaluation, motivating robots with similar physical interaction capabilities. Existing robotic hair-care and surface-following methods mainly rely on trajectory planning, compliance, force regulation, or tactile-conditioned policies, but deformable hair can remain in contact while gradually drifting across the end-effector, making local interaction difficult to regulate. We propose TacHair, a tactile contact-distribution guided online correction framework that represents high-resolution tactile observations as a spatial hair-contact distribution. A visuotactile imitation policy generates the nominal stroking motion, while a separately trained residual module corrects local contact deviations, separating task progression from contact recovery. We evaluate TacHair in 525 real-robot trials across five head geometries and three hair conditions. A successful stroke requires both sufficient task progression and contact maintenance; our method improves success from 42.9% to 62.3% and contact maintenance from 59.4% to 88.0% over the same visuotactile policy without correction. These results demonstrate spatial tactile contact distributions as an effective feedback representation for contact-preserving interaction with deformable and visually occluded surfaces. Demos, code, and datasets are available at https://tachair.github.io.
Self-Organization from Constrained Geometric Radiation
arXiv:2610.10621v1 Announce Type: new Abstract: How does dynamic order emerge spontaneously in closed systems without external driving? Existing paradigms all require external energy flows, temperature quenching, or slow driving. Here we report constraint-induced self-organization via geometric radiation in coupled metric evolution systems. Simulations reveal a universal four-stage cycle: stress accumulation, super-exponential radiation, chaotic collapse, and convergence to a fractal limit cycle, a novel attractor topology we term the wedge-shaped attractor, with five quantized curvature states and fractal micro-fluctuations. We identify four jointly sufficient conditions: an irreversible geometric horizon, persistent stress injection from quantum coherence, endogenous geometric tension between incompatible curvatures, and effective fluctuations. Their synergy triggers a critical avalanche at the horizon boundary. We prove three theorems: the Geometric Horizon Theorem, the Geometric Energy Dissipation Theorem (implying wave-like entropy evolution in closed systems), and the Radiation as Phase Transition Channel Theorem. We further establish the Constraint-Induced Self-Organization Theorem: these conditions guarantee the complete cycle with probability one. Systematic scans reveal a critical noise threshold and power-law scaling of radiation onset. We verify universality across 12 configurations, multiple noise types, and three geometric flows. This work establishes a new paradigm for closed-system self-organization, forging an exact mathematical duality between classical nonlinear constraints and gravitational horizons.
Toward Reliable Patient-Specific Aortic Strain Mapping from 4D CTA: Validation, Spectral Structure, and Clinical Potential
arXiv:2610.10913v1 Announce Type: new Abstract: Objective: To develop and validate a framework for spatially resolved ascending aortic strain estimation from 4D computed tomography angiography (CTA) and characterize reliability across segmentation sources and spatial scales. Methods: We constructed patient-specific aortic strain maps using nnU-Net segmentation, optimized surface remeshing, deformable registration, and mesh-based strain estimation. Finite element (FE)-derived synthetic 4D CT sequences provided controlled reference deformations for registration validation. Strain reliability was characterized across segmentation sources and Laplace-Beltrami (LB) spectral scales. Associations with diameter, aortic height index (AHI), and age were assessed. Results: Registration-derived median strain showed minimal bias relative to FE-reference strain for observer-derived (0.0013) and nnU-Net-derived meshes (0.0056). Median strain exhibited low bias and narrow 95% limits of agreement (LoA) across segmentation sources (interobserver: bias = -0.0045, LoA [-0.0490, 0.0399]; nnU-Net versus observer: bias = -0.0052, LoA [-0.0340, 0.0237]). Differences were greater for 95th-percentile strain, particularly in the interobserver comparison. LB spectral analysis showed preferential preservation of low-frequency strain organization, with k = 5 reconstructions reducing segmentation-related disagreement by 52-56%. The 95th-percentile areal strain was inversely associated with maximum ascending aortic diameter (r = -0.493, p = 0.005), AHI (r = -0.488, p = 0.005), and age (r = -0.479, p = 0.006). Three high-grade AR cases showed heterogeneous regional strain patterns. Conclusion: This framework reproducibly estimates ascending aortic strain from 4D CTA, characterizes its reliability across segmentation sources and spatial scales, and demonstrates its potential for noninvasive assessment of patient-specific aortic mechanics.
A General $\widetilde{\Omega}(\sqrt{T \gamma_T})$ Lower Bound for Kernel Bandits
arXiv:2610.11082v1 Announce Type: new Abstract: The kernel bandit problem consists of sequentially optimizing an unknown function with noisy feedback, where the function has bounded norm in a given Reproducing Kernel Hilbert Space (RKHS). A central quantity in the regret analysis of kernel bandits is the maximum information gain $\gamma_T$. In particular, the best existing upper bounds scale as $\sqrt{T\gamma_T}$ up to log factors, and nearly-matching lower bounds have been derived for specific kernels such as squared exponential and Mat\'ern. However, lower bounds for general kernels are lacking, thus making it unclear in what generality the upper bounds are near-optimal. In this paper, we establish a general $\Omega(\sqrt{T\gamma_T/\log T})$ minimax regret lower bound for non-constant continuous kernels on compact domains, establishing near-optimality (within log factors) in a very general sense. We show that the log factor appearing in this bound is unavoidable in general, but that it can be removed under certain conditions. Among other things, our findings imply that the minimax-optimal scaling is exactly $\Theta(\sqrt{T\gamma_T})$ (i.e., within constant factors) for the Mat\'ern-$\nu$ kernel with $\nu \in (0,2)$, $\gamma$-exponential kernel with $\gamma \in (0,2)$, and certain piecewise-polynomial kernels.
Nonlinear Feedback in Josephson Circuit Optimization: Application to a Kerr-Reversal JTWPA
arXiv:2610.10715v1 Announce Type: new Abstract: Optimizing Josephson-based nonlinear microwave devices is computationally demanding because the straightforward approach requires exploring broad circuit parameter spaces through expensive nonlinear simulations. Josephson Circuits Optimizer addresses this problem by using harmonic balance simulations in two stages: fast linear simulations to select promising circuit configurations according to properties such as impedance and phase matching, followed by nonlinear simulations to optimize their operating conditions. However, pump-induced effects such as impedance renormalization and Kerr-induced modifications of phase matching are not captured during the initial linear stage. We therefore introduce a nonlinear feedback extension that transfers information from the pumped response back to the linear optimization. The approach is investigated using a Josephson traveling-wave parametric amplifier with a reversed-Kerr architecture, whose model is validated against experimental gain measurements. We then evaluate the difference between the linear and pumped response values of a metric based on the input reflection coefficient, while excluding configurations with insufficient third-harmonic suppression. This discrepancy is considered a candidate feedback observable, as it exhibits features similar to the gain landscape, supporting its use in subsequent optimization cycles.
Omnisolver: An extensible interface to Ising spin-glass and QUBO solvers: adding a distributed GPU brute-force plugin
arXiv:2610.10542v1 Announce Type: new Abstract: This software update extends Omnisolver with \texttt{omnisolver-bruteforce}, a first-class plugin for exact exhaustive search of QUBO and Ising instances on CUDA-enabled GPUs. The update contributes three components: packaging of the single-GPU brute-force kernel of [Computer Physics Communications 260, 107728, 2021] as a first-class Omnisolver plugin with a uniform Python and CLI interface, a Ray-based distributed framework that splits the search into $2^{k}$ fixed-variable subproblems and dispatches them across multiple GPUs and hosts, with a controller that gathers and merges partial results, and numerical stabilization of the fast \texttt{float32} ground-state path through compensated incremental updates, periodic exact energy re-anchoring, and best-buffer refresh. Stabilization activates automatically for $N \geq 40$ leaving the public sampler API is unchanged. On dense random Ising instances we measure exhaustive solve times up to $N=60$ on $8\times$ NVIDIA H100 (96 GB) GPUs, reaching $\approx\!3.15$ days at $N=60$ in close agreement with the empirical $t_{N+1}=2\,t_N$ doubling rule, with the distributed sampler reaching the full $\approx\!8\times$ speedup over a single H100 from $N\!\gtrsim\!44$ onwards. The plugin thereby serves as a practical ground-truth oracle for state-of-the-art classical heuristics such as the Simulated Bifurcation Machine.
VICO: Visual Environments Co-Evolving for Vision-Language Model Reasoning
arXiv:2610.10782v1 Announce Type: new Abstract: Reinforcement learning with verifiable rewards (RLVR) has become a standard recipe for post-training vision-language models (VLMs), but it typically assumes a static training environment. As the actor improves, fixed tasks drift out of its learning frontier: many become trivial, others remain unsolvable; and the learning signal collapses. We argue that VLM post-training should evolve the visual environment alongside the actor, not just the actor itself. We propose VICO, a co-evolutionary framework in which an actor and an Environment-as-Rewriter (EnvRewriter) are trained jointly: the EnvRewriter edits verifiable image-side structures, such as scene graphs, chart tables, or protected region masks, and re-renders them to produce label-valid training samples whose difficulty is calibrated to the actor's current ability through a pass-rate-based reward. This loop continuously realigns task difficulty with actor capability without any additional human annotation. Across nine multimodal benchmarks spanning mathematical reasoning and visually grounded understanding, VICO-8B improves over its base model by up to +5.0% on out-of-domain tasks, surpasses the strongest self-evolution and text-editing co-evolution baselines by +4.3% and +8.4% respectively, and stays comparable to chart-specialized RLVR methods using 16-160 times fewer labeled samples. By shifting from human-labeled supervision to image-editing co-evolution, VICO offers a scalable path beyond static-corpus RLVR for visual reasoning.
Teaching a Robot Dog New Tricks: Diverse Quadruped Skills via Combined Reinforcement and Imitation Learning with Adversarial Task Selection
arXiv:2610.10601v1 Announce Type: new Abstract: Reinforcement Learning (RL) has enabled legged robots to perform a range of skills in single-task settings. However, applications such as farm robotics or space exploration require diverse skills such as locomotion, digging, or close-range surveying. Training an end-to-end policy to address this problem remains difficult due to challenges such as sample inefficiency and gradient conflict between tasks in multi-task learning. We propose a three-stage method that trains a single policy to perform distinct tasks such as walking, digging, and hopping, and compose them into novel behaviors such as crawling. First, multiple teacher policies are trained using RL on narrowly defined tasks. Then, two additional stages train a student policy with a multi-teacher distillation setup that uses a combined RL and Imitation Learning (IL) objective under an adversarial task selection process that focuses training on the worst-performing task. With this method, we train a student policy that performs 22 tasks using 8 teachers. Evaluations show our method preserves motion quality and tracks commands more accurately than PPO and distill-then-finetune baselines, and in some cases generalizes to new tasks without explicit training. Finally, we demonstrate real-world robustness by deploying the resulting policy on a Unitree B1 quadruped. Video: https://youtu.be/V9yX04EBcFA
Combining Error Verification and Privacy Amplification in Quantum Key Distribution
arXiv:2610.10702v1 Announce Type: new Abstract: Quantum Key Distribution (QKD) protocols typically involve two separate steps of error-verification and privacy amplification. Both steps involve the application of a hash function sampled from a suitable hash family to classical data. A potentially simpler alternative is to implement a single-step hashing procedure, announce and compare a fixed number of bits of its output for error-verification, and retain the remaining bits as the final output key. This reduces the complexity of classical postprocessing, the amount of classical communication, and the amount of randomness needed in the protocol. While this construction has appeared in the literature, a suitable mathematical justification is lacking. In this work, we show that the security of such protocols, under certain simple conditions, follows from a slightly modified notion of secrecy of the protocol without the final error-verification step. In particular, we show that for most entropy-based proofs, the combined protocol can be implemented with no change to the key rate, and a minor change to the security parameter.
Quantum Networking at the Speed of Quantum Computation: Kilohertz Entanglement in a Heterogeneous Quantum System
arXiv:2610.10705v1 Announce Type: new Abstract: Distributed quantum computing requires rapid, high-fidelity entanglement between qubits in separate processor modules. Photonic links between single-atom qubits provide switchable, long-range connections, but photon loss limits their entanglement rate. Here we entangle a trapped $^{138}$Ba$^{+}$ ion with a silicon-vacancy (SiV$^{-}$) center in a diamond nanophotonic cavity. A photon emitted by the ion is converted in wavelength and encoding, reflected from the SiV$^{-}$--cavity system and detected, so that a single detected photon heralds entanglement. Including all conversion losses, we generate ion--SiV$^{-}$ Bell pairs at an average rate of 1.03(1)\,kHz with a fidelity of 87.9(7)\%. This is four times the rate of the fastest photonic link between two trapped ions and corresponds to one Bell pair per millisecond, matching the cycle time of planned trapped-ion processors.
From Investigation Failures to Reliable SOC Agents: Understanding and Improving LLM-Based Alert Triage
arXiv:2610.10608v1 Announce Type: cross Abstract: Security operations centers (SOCs) must triage large volumes of alerts, most of which are benign, while missed attacks can remain uninvestigated. Tool-using large language model (LLM) agents can retrieve evidence during triage, but it remains unclear how reasoning strategies determine what to gather and when an investigation is sufficient to close an alert. We study five representative approaches spanning single-pass tool use, iterative retrieval, sampled investigations, self-review, and explicit verification. To support this study, we build ALERT-BENCH, an interactive benchmark that replays enterprise telemetry through a live SIEM and requires each system to retrieve evidence. Across 1,247 alerts from a multi-stage attack scenario, every approach missed at least 40.4% of attack-related alerts. Trace analysis shows that attack alerts are more likely to be dismissed when searches return no records, same-context review has negative net correction, and dismissal receives no consistently stronger investigation than escalation. Based on these findings, we further design AIDA (Adversarial Investigation and Dialectical Analysis), a multi-agent framework that requires an explicit proposed decision before independent challenge and stronger evidentiary requirements before dismissal. AIDA preserves investigation history in an append-only Investigation Ledger and keeps the challenge in a separate reasoning context. A separate Judge adjudicates the proposed decision and challenge against evidence, resolving the alert or requesting another round when evidence is missing. On the same alerts, AIDA achieves an F1 score of 0.958, compared with 0.371-0.744 for the studied approaches, and reduces the false-negative rate from 40.4% to 3.1% while escalating 18.4% of alerts to analysts. These results show that structuring evidence retrieval and decision review can substantially improve agentic SOC triage.
Wieszcz-XIX: A 3.1-Billion-Word Corpus of Pre-1918 Polish and Temporally Bounded Language Models Trained From Scratch
arXiv:2610.10592v1 Announce Type: new Abstract: Historical Polish is well documented as a language but annotated in machine-readable form only to about a million words for the period this paper covers; the rest sits behind optical character recognition of variable quality. We present Wieszcz-XIX, a corpus of 6.75 billion tokens (about 3.1 billion words) in 294,369 documents, most of them periodical issues, of Polish published from 1800 to 1918, assembled from Wolne Lektury and the Internet Archive by a pipeline that filters, deduplicates, audits for post-1918 leakage and splits at the document level. It is over three orders of magnitude larger than the annotated corpus of the same period, and we quantify its defects: recognition corruption against a false-positive floor, near-identical duplication, which is removed, and post-1918 leakage, which is excluded from the training corpus itself down to a known residue of 0.04 to 0.38% of its bytes, found in the transcribed source, so the published corpus is the trained one document for document. On a hand-corrected sample the character error rate is 0.68% where the text is legible, and 45% of the sampled passages cannot be corrected. On it we train a ladder of decoder-only models from 47M to 349M parameters from scratch, and measure their temporal boundedness. Against two modern Polish base models, one far larger, the 349M shows a crossover, as does the 107M against the comparator of its size: post-1918 vocabulary costs them about 3.1 bits per byte more than period vocabulary, a gap the comparators do not show, and period vocabulary costs them fewer bits than it costs the comparators. Shown period text, the models keep its spelling and the comparators only partly. Adding parameters gains about twice as much as a second pass over the data. We release the corpus, code and weights. Content warning: the models reproduce period prejudice, including antisemitic statements.
Diffu-LoRA: A Novel Low-Rank Adaptation for Personalized Diffusion Models
arXiv:2610.10550v1 Announce Type: new Abstract: Personalizing text-to-image diffusion models from a few reference images requires preserving subject identity while following prompts that describe new contexts. Full-model fine-tuning is parameter-intensive, whereas low-rank adaptation (LoRA) reduces the number of trainable parameters but leaves open how adaptation capacity should be distributed across layers. We introduce Diffu-LoRA, a parameter-efficient method that learns this allocation through gated low-rank adaptation. Diffu-LoRA inserts trainable low-rank components into the linear layers of Transformer blocks and assigns a learnable gate to each component. Bilevel optimization updates the adaptation weights and gate parameters on separate data splits, while progressive pruning removes components with the lowest gate values to meet a prescribed rank budget. This procedure allocates adaptation capacity nonuniformly across layers while keeping the pretrained backbone frozen. Experiments with Stable Diffusion on subjects from DreamBooth and additional collected datasets show improved overall subject fidelity and prompt alignment relative to the evaluated fine-tuning baselines. Ablation studies examine the contributions of bilevel optimization, progressive pruning, and adapter placement. These results support learned rank allocation as a practical approach to parameter-efficient diffusion model personalization.
Seeing Through the Glare: A Multi-Source Benchmark and Ocular-Adaptive Pixel MeanFlow for Eyeglass Reflection Removal
arXiv:2610.10703v1 Announce Type: new Abstract: Eyeglass reflection removal is important across smartphone imaging, video conferencing, and other face-centric visual applications. The task is challenging because reflections range from mild photometric contamination to severe ocular occlusion, requiring selective correction and plausible reconstruction without altering identity or natural appearance. Existing datasets cover limited reflection conditions, constraining generalization to complex real-world scenes and systematic evaluation. We introduce \textbf{OcuBench}, a multi-source benchmark comprising 10,280 controllable synthetic pairs, 732 real-input pseudo-pairs, and 458 independent real-world test images, supporting both paired evaluation and assessment beyond generated supervision. We further propose \textbf{OcuFlow}, an ocular-adaptive pixel MeanFlow (pMF) framework for efficient, detail-preserving restoration. It combines geometry-adaptive representation with one-step pMF to focus reconstruction on reflection-obscured ocular regions, together with native-resolution frequency-preserving synthesis to retain reliable observed details. Experiments across diverse reflection conditions demonstrate that OcuFlow achieves consistent advantages in reflection removal quality, ocular fidelity, and efficiency. In a blind user study, it receives $67.32\%$ of selections, $6.2\times$ the next-best share. Both the code and dataset will be released.