Search arXiv⌕ Search

SEARCH · Search arXiv

Search Search arXiv

Search indexed arXiv papers on artificial intelligence, large language models, computer vision and robotics. Read source abstracts and follow links to arXiv.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 1,783 records · Page 99Linked to original sources
arXiv2026

Dynamics of Interfaces in the Two-Dimensional Wave-Pinning Model

We study the mass-conserved reaction-diffusion system known as the wave-pinning model, which serves as a minimal framework for describing cell polarity. In this model, the interplay between reaction kinetics and slow diffusion forms a sharp interface that partitions the domain into high- and low-concentration regions. We perform a detailed asymptotic analysis and derive higher-order approximation equations governing the motion of this interface. Our results show that on a fast timescale, the interface evolves via propagating front dynamics, whereas on a slow timescale, it evolves as an area-preserving mean curvature flow. Furthermore, using the derived free boundary problem, we demonstrate that on a significantly slower timescale, an interface whose endpoints lie on the domain boundary drifts along the boundary toward regions of higher curvature. In summary, our analysis reveals that the interface dynamics in the wave-pinning model exhibit a hierarchy of three distinct timescales: wave propagation on a fast timescale, curvature-driven area-preserving evolution on a slow timescale, and motion along the boundary on a significantly slower timescale.

math.DS↗
arXiv2026

Pedagogical Friction in Password Creation

Users often make security- and privacy-relevant decisions without a clear understanding of the rules that govern safe behavior. We introduce pedagogical friction, a design approach that inserts brief, instructional interactions at the moment of action. We evaluate this approach in the context of password creation, a familiar task with clear quality criteria. We conducted a randomized study with 128 participants across four interface conditions that varied the depth and interactivity of guidance. We assessed three outcomes: (1) rule compliance in a subsequent password task without guidance, (2) accuracy on survey questions tied to password rules, and (3) behavior-knowledge alignment, which captures whether participants who correctly followed a rule also recognized it on the survey. Across the guided conditions, participants corrected most rule violations in the follow-up task and showed high behavior-knowledge alignment. Survey results suggested clearer advantages for some rule types, especially symbol related questions. These results position pedagogical friction as a lightweight intervention for security- and privacy-critical interfaces.

cs.HC↗
arXiv2026

PsyCLIENT: Client Simulation via Conversational Trajectory Modeling for Trainee Practice and Model Evaluation in Mental Health Counseling

LLM-based client simulation provides a scalable approach to novice counselor training, counseling-dialogue synthesis, and interactive evaluation of automated counseling systems. However, existing approaches are limited by insufficient profile diversity, weak behavioral grounding, and the lack of open Chinese-language resources for simulated counseling clients. We propose PsyCLIENT, a framework that conditions simulated-client responses on client profiles, dialogue histories, and conversational trajectories specifying target behaviors and content constraints at each client turn. We also construct PsyCLIENT-CP, a dataset of 120 Chinese client profiles spanning 60 counseling topics. Evaluations involving 24 professional counselors show that PsyCLIENT receives higher ratings of perceived authenticity and training utility than the comparison methods. In a separate source-identification study, PsyCLIENT dialogues were more frequently misclassified as human-client interactions than those generated by the baselines. These results suggest that conversational trajectory modeling can transform static client profiles into behavior-guided, dynamically unfolding counseling interactions.

cs.CL↗
arXiv2026

A unified framework for data-driven Bell inequalities

Although the original EPR paradox was formulated in terms of position and momentum, most studies of these phenomena have focused on measurement scenarios with only a discrete number of possible measurement outcomes. Here, we present a framework for studying non-locality that is agnostic to the dimension of the physical systems involved, allowing us to probe purely continuous-variable, discrete-variable, and hybrid continuous/discrete non-locality. Our approach allows us to find the optimal Bell inequality for any given measurement scenario and quantifies the amount of non-locality that is present in measurement statistics. This formalism unifies the existing literature on continuous-variable non-locality and allows us to identify new states in which Bell non-locality can be probed through homodyne detection. Notably, we find the first example of two-mode continuous-variable non-locality that cannot be mapped to a binary Bell inequality. Moreover, we provide several examples of simple hybrid DV-CV entangled states that could lead to near-term violation of Bell inequalities.

quant-ph↗
arXiv2026

An efficient mixed-integer linear programming formulation for solving influence diagrams

Influence diagrams represent decision-making problems with interdependencies between random events, decisions, and consequences. Traditionally, they have been solved using tailored algorithms that determine the expected utility-maximizing decision strategy. Towards this end, most recent state-of-the-art solution approaches convert them into mixed-integer linear programming (MILP) models that can be solved using off-the-shelf MILP solvers. Computationally, these approaches are most efficient for influence diagrams that represent periodic (or sequential) decision processes, which can be cast as partially observable Markov Decision Processes. However, they are inefficient for problems that lack this periodic structure or whose influence diagrams have large state spaces. The efficient MILP formulation presented in this paper is specifically designed for influence diagrams that are challenging for earlier MILP formulations. Additionally, this formulation can be adapted to optimize conditional value-at-risk and to incorporate chance and logical constraints, thereby retaining the modeling flexibility of MILP-based methods. Finally, computational experiments on a range of problems from the literature are reported to show that the proposed formulation is much more efficient than earlier MILP formulations, particularly in problems which cannot be modeled as partially observable Markov decision processes and which lead to influence diagrams whose decision or chance nodes may have large state spaces.

math.OC↗
arXiv2026

Decoding Weak Phase Amplitudes in Charmed Baryon Decays

In this Letter, we present a strategy to decode decay amplitudes with different weak phases in two-body charmed baryon decays, addressing a long-standing challenge to understanding CP violation in this sector. Using SU(3) flavor symmetry, we disentangle the amplitudes defined through $M=λ_s A_s + λ_b A_b$ ($λ_i = V_{ui}V_{ci}^*$), enabling a data-driven determination of their relative hierarchy across Cabibbo-favored, singly, and doubly Cabibbo-suppressed modes. Applying this framework to current data, we find that the ratio $A_b/A_s$ is approximately 100 times larger than expectation of $\mathcal{O}(1)$ at $2.1σ$ level. This unexpected pattern may be accommodated by SU(3) breaking effects, but it can also point to new physics. We further show that the Lee-Yang parameters provide an independent and complementary probe. To resolve this ambiguity in the origin of the enhancement, we propose a decisive discriminator using decay modes insensitive to large SU(3) breaking, in particular $Ξ_c^0 \to pK^-$ and $Ξ_c^0 \to Σ^+π^-$ as golden channels. Measurements of these modes will distinguish between symmetry breaking effects and new physics, establishing a clean pathway to probe CP violation in the charm baryon sector.

hep-ph↗
arXiv2026

Noise-Robust Quantum-Inspired Synthetic Minority Oversampling for Condition Monitoring and Fault Classification in Industrial and Energy Systems

Class imbalance remains a major challenge in industrial condition monitoring and fault classification, where minority fault patterns can be poorly represented during machine learning. This work develops a noise-aware quantum-inspired oversampling framework and systematically evaluates its effectiveness and robustness for industrial and energy-system fault diagnosis. The principal artificial-intelligence contribution is a noise-aware extension of the Quantum Synthetic Minority Oversampling Technique (QSMOTE), in which quantum-inspired perturbations are introduced directly into the compact-swap-test-based similarity estimation used during synthetic minority-sample generation. Consequently, noise propagates through overlap estimation and angle computation to the generated samples, enabling direct assessment of the robustness of the oversampling mechanism rather than perturbing only the downstream classifier. The engineering application is evaluated on four public datasets covering solar-panel faults, bearing condition monitoring, engine failures, and industrial fault detection. Five classical classifiers are used to compare performance on the original imbalanced data, after QSMOTE-based balancing, and under noisy-QSMOTE. The results show substantial classifier-dependent benefits from QSMOTE, including improvements of up to 170% on the engine-failure dataset and accuracy exceeding 0.99 on the industrial fault-detection dataset. Robustness experiments across six quantum-inspired noise channels further demonstrate that perturbations in similarity estimation can produce markedly different downstream effects depending on the noise model and classifier. The proposed framework therefore provides a reproducible approach for jointly assessing class-imbalance mitigation and noise robustness of quantum-inspired oversampling in practical industrial artificial-intelligence applications.

quant-ph↗
arXiv2026

Survival is the Only Reward: Sustainable Self-Training Through Environment-Mediated Selection

Self-training systems often degenerate due to the lack of an external criterion for judging data quality, leading to reward hacking and semantic drift. This paper provides a proof-of-concept system architecture for stable self-training under sparse external feedback and bounded memory, and empirically characterises its learning dynamics and failure modes. We introduce a self-training architecture in which learning is mediated exclusively by environmental viability, rather than by reward, objective functions, or externally defined fitness criteria. Candidate behaviours are executed under real resource constraints, and only those whose environmental effects both persist and preserve the possibility of future interaction are propagated. The environment does not provide semantic feedback, dense rewards, or task-specific supervision; selection operates solely through differential survival of behaviours as world-altering events, making proxy optimisation impossible and rendering reward-hacking evolutionarily unstable. Analysis of semantic dynamics shows that improvement arises primarily through the persistence of effective and repeatable strategies under a regime of consolidation and pruning, a paradigm we refer to as negative-space learning (NSL), and that models develop meta-learning strategies (such as deliberate experimental failure in order to elicit informative error messages) without explicit instruction. This work establishes that environment-grounded selection enables sustainable open-ended self-improvement, offering a viable path toward more robust and generalisable autonomous systems without reliance on human-curated data or complex reward shaping.

cs.AI↗
arXiv2026

TaxoBench: Can Deep Research Agents Retrieve and Organize? Evaluating the Synthesis Gap with Expert Taxonomies

Deep Research Agents increasingly automate survey writing, yet existing benchmarks do not jointly test whether they retrieve the papers experts consider essential and organize those papers into paper-grounded taxonomies. We introduce TaxoBench, a benchmark built from 72 highly cited LLM surveys, 3,815 cited papers, and their expert-authored taxonomies. TaxoBench evaluates systems in two settings: Deep Research mode measures end-to-end retrieval and organization from a topic, while Bottom-Up mode provides the expert paper set and isolates organization. We evaluate leaf-level assignments with ARI and V-Measure and hierarchy-level structure with US-TED, US-NTED, and Sem-Path. Across 7 Deep Research Agents and 16 LLM configurations, the best agent retrieves only 20.92% of expert-cited papers, and none of 70 standard Bottom-Up runs reaches the experts' average depth of 4.86. A controlled probe shows that models which match this depth do so by fragmenting the taxonomy, reducing alignment with the expert reference. We further find that raw Sem-Path remains near a no-organization floor even when a newer model generation gains 3.68 pp ARI; after depth matching, humans lead on all 10 matched surveys by 13.27 pp. These results identify retrieval and hierarchical organization as separate bottlenecks and show why hierarchy metrics must be calibrated before they are used to compare models.

cs.CL↗
arXiv2026

On the Runway Cascade of Transformers for Language Modeling

In decoder-only (causal) transformers, the computation graph created by causal masking routes information through both direct-path attention and indirect paths formed by intermediate tokens. We denote these indirect paths between token pairs as their runways. We argue that certain failure modes of causal transformers as observed by a growing body of recent works are likely exacerbated by a misalignment between these two information propagation modes. We formalize runway cascade as a phenomenon whereby this misalignment results in redundancies and irrelevant information cascading to token representations despite adequately learned attention patterns. As a solution, we propose runway-aware rewiring as a more explicit way of incorporating runway context directly into each token's direct-path attention. This mechanism re-wires the attention pattern for each token based on a summary of its runway landscape, enabling awareness of accumulating representational influences and allowing for more balanced information propagation. Our proposed methodology introduces no additional parameters and can seamlessly be integrated into standard attention mechanism. Empirically, our rewired transformer results in steady improvements in general language modeling as well as noticeably stronger information retrieval and extrapolation abilities compared to standard transformers.

cs.LG↗
arXiv2026

On isoperimetric local-Bollobás-Thomason inequalities

We prove the following isoperimetric-type inequality: for every convex body $K$ in $\mathbb R^n$ and some $σ\subset[n]:=\{1,\dots,n\}$ there exists a suitable Hanner polytope $B_K$ with the same volume as $K$ and such that the volume of each of its orthogonal projections onto every subspace whose basis is formed by the canonical vectors $\{e_i:i\inτ\cup([n]\setminusσ)\}$, for every $τ\subseteqσ$, bounds from below the volume of the corresponding projections of $K$.

math.MG↗
arXiv2026

Construction and Decoding of Convolutional Codes with optimal Column Distances

The construction of Maximum Distance Profile (MDP) convolutional codes in general requires the use of very large finite fields. In contrast convolutional codes with optimal column distances maximize the column distances for a given arbitrary finite field. In this paper, we present a construction of such convolutional codes. In addition, we prove that for the considered parameters the codes that we constructed are the only ones achieving optimal column distances. The structure of the presented convolutional codes with optimal column distances is strongly related to first order Reed-Muller block codes and we leverage this fact to develop a reduced complexity version of the Viterbi algorithm for these codes.

cs.IT↗
arXiv2026

TabClustPFN: A Prior-data Fitted Network for Tabular Data Clustering

Prior-data Fitted Networks (PFNs) have reframed supervised tabular learning as single-pass in-context inference without per-dataset optimization. Extending this paradigm to clustering is appealing yet fundamentally more challenging, due to absent supervision, unknown cluster cardinality, and label switching inherent to partition outputs. Existing PFN-based clustering methods address these challenges only partially, either requiring known cardinality as input or relying on unstable label-ordering conventions and overly restrictive synthetic priors. We introduce TabClustPFN, a clustering PFN that resolves all challenges jointly through co-designed prior, objective, and architecture. Our hybrid pretraining prior captures heterogeneous real-tabular geometry; our decoupled partition inference network--cardinality inference network architecture jointly infers cluster assignments and cardinality in a single pass; and our SoftARI training objective is permutation-invariant by construction, eliminating the need for any label-ordering convention. On a 44 curated real-world tabular benchmark, TabClustPFN achieves state-of-the-art clustering performance against classical, deep, and amortized baselines, with runtime comparable to efficient baselines. Code is available at https://github.com/Tianqi-Zhao/TabClustPFN.

cs.LG↗
arXiv2026

Linear systems on blow-ups of Hirzebruch surfaces

Motivated by various equivalent versions of the SHGH conjecture for $\mathbb P^2$ blown up at very general points, we propose a similar conjecture for Hirzebruch surfaces. We prove that this conjecture is true for the Hirzebruch surface $\mathbb F_e$ blown up at $r\leqslant e+5$ very general points.

math.AG↗
arXiv2026

Structural properties of distance-bounded phylogenetic reconciliation

Phylogenetic reconciliation seeks to explain host-symbiont co-evolution by mapping parasite trees onto host trees through events such as cospeciation, duplication, host switching, and loss. Finding an optimal reconciliation that ensures time-consistency is computationally hard when timing information is incomplete, and the complexity remains open when host switches are restricted by a fixed maximum distance $d$. While the case $d = 2$ is known to be polynomial, larger values are unresolved. In this paper, we study the cases $d = 3$ and $d = 4$. We show that although arbitrarily large cycles may occur, it suffices to check only bounded-size cycles (we provide a complete list), provided the reconciliation satisfies acyclicity (i.e., time-consistency) in a stronger sense, which we call strong acyclicity. Strong acyclicity is associated with incorporating discrete time slices into the model.

q-bio.PE↗
arXiv2026

Compressed Set Representations based on Set Difference

We introduce a compressed representation of sets of sets that exploits how similar they are to each other. Our representation supports access, membership, predecessor and successor queries on the sets within logarithmic time. It also computes the support of any element (i.e., the sets containing it) in optimal time. In addition, we give a new MST-based construction algorithm for the representation that outperforms standard ones.

cs.DS↗
arXiv2026

The Cofinality of Generating Familes

The topology of a separable metrizable space $M$ is \emph{generated} by a family $\mathcal{C}$ of its subsets provided that a set $A\subseteq M$ is closed in $M$ if and only if $A\cap C$ is closed in $C$ for each $C\in \mathcal{C}$. The \emph{sequentiality number}, $\mathop{seq}(M)$, and \emph{$k$-ness number}, $\mathop{k}(M)$, of $M$, are the minimum size of a generating family of convergent sequences, respectively compact subsets. Let $\mathfrak{b}$ be the minimum size of an unbounded set in $ω^ω$ with the mod finite order. For a cardinal $κ$, the \emph{covering number}, $\mathop{cov}(κ)$, is the minimum size of a family of countable subsets of $κ$ so that every countable subset of $κ$ is contained in an element of the family. It is shown using the Tukey order on relations that (1) $\mathop{seq}(M)=\mathop{cov}(|M|)\cdot \mathfrak{b}$, unless $M$ is locally small (every point of $M$ has a neighborhood of size strictly less than $|M|$) in which case $\mathop{seq}(M)=\lim_{μ<|M|} \mathop{cov}(μ)\cdot \mathfrak{b}$ and (2) $k(M)$ is in the interval $[kc(M)\cdot\mathfrak{b},\mathop{cov}(kc(M))\cdot \mathfrak{b}]$, where $kc(M)$ is the minimum number of compact sets that cover $M$. Solutions to problems of van Douwen's on the $k$-ness number of analytic and of co-analytic spaces are deduced.

math.LO↗
arXiv2026

GFlowPO: Generative Flow Network as a Language Model Prompt Optimizer

Finding effective prompts for language models (LMs) is critical yet notoriously difficult: the prompt space is combinatorially large, rewards are sparse due to expensive target-LM evaluation. Yet, existing RL-based prompt optimizers often rely on on-policy updates and a meta-prompt sampled from a fixed distribution, leading to poor sample efficiency. We propose GFlowPO, a probabilistic prompt optimization framework that casts prompt search as a posterior inference problem over latent prompts regularized by a meta-prompted reference-LM prior. In the first step, we fine-tune a lightweight prompt-LM with an off-policy Generative Flow Network (GFlowNet) objective, using a replay-based training policy that reuses past prompt evaluations to enable sample-efficient exploration. In the second step, we introduce Dynamic Memory Update (DMU), a training-free mechanism that updates the meta-prompt by injecting both (i) diverse prompts from a replay buffer and (ii) top-performing prompts from a small priority queue, thereby progressively concentrating the search process on high-reward regions. Across few-shot text classification, instruction induction benchmarks, and question answering tasks, GFlowPO consistently outperforms recent discrete prompt optimization baselines.

cs.AI↗