arXiv:2605.15789v1 Announce Type: new Abstract: Uncertainty quantification is essential in safety-critical settings--from autonomous driving to aviation, finance, and health--where decisions must rely on conservative bounds rather than point estimates. Predictor-level intervals (e.g., from quantile regression, conformal prediction, variance networks, or Bayesian models) generally do not compose: adding two per-variable intervals need not yield a valid interval for their sum or preserve coverage. In aviation, Gaussian overbounding replaces complex error distributions with a conservative Gaussian whose tails dominate the truth, so conservatism propagates through linear operations. Yet classical overbounds are global, often overly conservative, and hard to adapt to feature-conditioned errors. We propose a unified learning framework that trains neural networks to produce context-aware Gaussian overbounds--mean and scale--with provable conservatism on a finite quantile grid and, under three explicit regularity assumptions, continuous-tail conservatism on a certified interval. Our overbounding loss enforces conservativeness at selected quantiles while penalizing distributional distance with a Wasserstein-style term. The learned bounds support conservative linear-combination and convolution analysis on the enforced grid, and on the certified interval when assumptions hold, while being less redundant than traditional methods. We provide a scoped analysis of discrete-to-continuous conservatism and compact-domain objective regularity, and validate on synthetic data and real-world datasets, including multipath, ionospheric, and tropospheric residual errors. Across these settings, the method yields tighter bounds while maintaining conservatism on the enforced grid and in experiments. The framework is modality-agnostic and applicable to learning systems that require conservative, feature-conditioned uncertainty estimates in dynamic environments.
Science Journals
arXiv:2605.15790v1 Announce Type: new Abstract: Retrieval-Augmented Generation (RAG) improves reliability of large language models by incorporating external knowledge, but the retrieval process can introduce bias that propagates to generated outputs. This issue is particularly challenging in top-k settings, where multiple documents jointly influence generation. We propose a fairness-aware retrieval framework that models and controls this bias. Our approach combines controlled bias injection via reranking, a position-aware model of bias propagation, and an optimization formulation that balances relevance and fairness. We further introduce a scalable solution based on Quadratic Fairness via Dual Hyperplane Approximation (FARO), which enables efficient optimization through problem decomposition. Experimental results show that our method effectively mitigates generation bias while preserving relevance. This work provides a principled approach for fairness-aware retrieval in RAG systems.
arXiv:2605.15282v1 Announce Type: new Abstract: Literary translation requires balancing target-language fluency with faithfulness to the source. Recent large language models (LLMs) often produce fluent translations, but it remains unclear whether fluency corresponds to semantic preservation in literary text. We examine this relationship using 130,486 translated paragraphs from 106 novels in 16 source languages, including human, Google Translate, and TranslateGemma translations. Fluency is measured as original-likeness with a translationese classifier trained on paragraph part-of-speech n-grams, and faithfulness with the automatic translation evaluation metric COMET-KIWI. We control for paragraph length and find a consistent negative correlation between fluency and faithfulness. The pattern appears for both human and Google Translate, but is weaker and often non-significant for TranslateGemma. These results show that segment length matters for automatic evaluation and suggest a tradeoff between fluency and faithfulness in literary translation.
arXiv:2601.17198v3 Announce Type: replace Abstract: This paper proposes sufficient, yet more general conditions for applying FastTwoSum as an error-free transformation (EFT) under all faithful rounding modes. Additionally, it also identifies guarantees tailored to round-to-odd for establishing FastTwoSum as an EFT. This paper also describes a floating-point splitting tailored for round-to-odd that is an EFT where the distribution of bits is configurable (i.e., ExtractScalar for round-to-odd). Our sufficient conditions are more general than those previously known in the literature (i.e., it applies to a wider operand domain).
arXiv:2605.16001v1 Announce Type: new Abstract: A broadcast on a connected graph is a function f that assigns each vertex v an integer f(v) with 0 <= f(v) <= ecc(v) where ecc(v) denotes the eccentricity of v. A vertex u hears a broadcasting vertex v (with f(v)>0) if u is at distance at most f(v) from v. Beyond the classical broadcast domination problem, where every vertex is required to hear at least one vertex, two variants raise intriguing combinatorial and algorithmic questions. In an independent broadcast, no broadcasting vertex hears another broadcasting vertex, while a broadcast packing requires that every vertex hears at most one broadcasting vertex. The corresponding problems Broadcast Independence and Broadcast Packing ask for broadcasts of values at least k under these constraints, where the value is the sum of the broadcast values. We initiate a systematic study of the parameterized complexity of such problems. We prove that Broadcast Independence and Broadcast Packing are FPT parameterized by the treewidth plus the diameter of G, with a family of dynamic-programming algorithms over nice tree decompositions. We obtain as a corollary that both problems are FPT parameterized by k and the treewidth of G and XP for treewidth only. The latter result shows that the known algorithm for trees (Bessy and Rautenbach, DAM 2022) can indeed be extended to bounded treewidth graphs. On the negative side, we show that Broadcast Independence is W[1]-hard parameterized by the pathwidth of G. Note that this result completes the picture for parameter k and treewidth for Broadcast Independence since it is known to be W[1]-hard for k only. We complement these results by showing that a weighted version of both problems, where the input comes with a weight function on the edges, is W[1]-hard parameterized by the vertex cover of G. Finally, we provide a constant-factor approximation algorithm parameterized by treewidth for Broadcast Independence.
arXiv:2605.15281v1 Announce Type: new Abstract: Modern web test suites rot. A UI refactor breaks locators, a timing change causes race conditions, and within weeks developers abandon the suite entirely. This paper presents an AI-driven autonomous testing framework that addresses these failure modes through five integrated strategies - navigation reliability, context-aware selector generation, post-generation validation, smart wait injection, and failure learning - implemented over a containerised worker architecture that decouples orchestration from long-running browser execution. Evaluated across four production applications and 176 scenarios, the framework improves script generation success from 55% to 93%, achieves an 8x reduction in navigation failures, eliminates 80% of timing-related race conditions, and reduces test creation time by 75% compared to manual Selenium authoring. The framework extends naturally to security validation: testers describe attack scenarios in plain English - "try accessing another user's invoice" - which the agent converts to OWASP Top 10-aligned browser probes, detecting 85% of authentication bypass vulnerabilities and 95% of input validation flaws with false positive rates below 12%. Natural-language-driven security testing of this kind represents, to our knowledge, a novel contribution to the field.
arXiv:2605.15279v1 Announce Type: new Abstract: Piezoelectric resonators (PRs) have been seen as a competitive alternative to magnetic components. In PR-based converters, the motional current (in the LC series branch of the equivalent circuit) is vital for control proposes but cannot be measured directly. The difficulties to detect the zero-crossing points or to measure the amplitude of the motional current has been one of the most dominating obstacles that complicates the control strategies and limits the frequency range of the PR-based converters. This work discusses a ring-dot shaped piezoelectric transformer (PT) based motional current sensing method that provides current information with low-delay, low-loss and intrinsic isolation. It is physically proven that the proposed method is robust with various non-ideal factors of the piezoceramic and circuit implementation. Based on this, an event-driven control strategy is introduced, consisting of only a finite state machine, a PI loop, a low-speed ADC and several comparators. Experiments on a step-down PR-based converter verify that the proposed approach realize ZVS for all transitions within a switching cycle with reduced hardware and software resources, enhances stability and is capable of self-startup.
arXiv:2601.15637v2 Announce Type: replace Abstract: Elastoinertial turbulence (EIT) is a self-sustaining chaotic state resulting from the interplay between inertia and elasticity in the flow of dilute polymeric solutions, and its emergence is believed to limit the achievable drag reduction in turbulence flow using polymer additives. In the present study, we introduce a viscoelastic variant of spectral proper orthogonal decomposition (VESPOD) that decomposes velocity and polymeric stress fields of EIT together into well-defined orthogonal oscillating modes such that the decomposition is optimal in the terms of the total mechanical energy of the flow. Using this technique, we investigate the dominant coherently evolving structures underlying the dynamics of EIT in axisymmetric pipe flow. By analyzing distinct peaks in the leading eigenvalue of the VESPOD eigenvalue spectrum, we find that the dynamics of EIT in pipe flow is dominated by three distinct families of traveling waves, where the higher wavenumber structures of each family are simple harmonics of their respective fundamental waves. The radial velocity fields of the traveling waves are characterized by the formation of large-scale structures spanning the pipe radial direction. However, the polymeric stress fields corresponding to them are characterized by the formation of thin inclined sheets of high stress fluctuations at the critical layers of the respective waves, i.e.~ the locations where the wave speed of the VESPOD mode matches the mean streamwise velocity. Additionally, these sheets exhibit nested structures, where the polymeric sheets of faster waves are confined by those of the immediately slower waves.
arXiv:2605.15257v1 Announce Type: new Abstract: Chain-of-thought (CoT) monitoring is one of the most promising tools we have for detecting model misbehavior, but its effectiveness depends on models faithfully externalizing their reasoning. Motivated by this vulnerability, we study whether monitor-aware models are capable of obfuscating their reasoning to evade detection. We use synthetic document finetuning to expose eight models to realistic pre-training-style documents describing a CoT monitor and find that monitor-aware models consistently achieve higher rates of undetected misbehavior compared to unaware controls. This effect is weaker but still present on a harder agentic task. We also show that CoT controllability, a model's ability to reshape its own reasoning trace under an imposed constraint, is closely correlated with obfuscation success across the eight models studied ($r=0.800$, $p=0.017$). Monitor-aware models placed under equal reinforcement learning optimization pressure also learn to reward-hack without triggering a CoT monitor substantially faster than unaware controls. Together, these results suggest that knowledge of monitoring combined with high CoT controllability poses a risk to CoT-based monitoring.
arXiv:2605.15792v1 Announce Type: new Abstract: The long-standing goal of multimodal AI is to build unified models in which visual understanding and visual generation mutually enhance one another. Despite recent works such as BAGEL, BLIP3o achieves remarkable progress; In practice, however, this unification remains one-directional: understanding routinely guides generation, yet how and why generation can support understanding is rarely investigated. We revisit this asymmetry and propose Generation-to-Understanding (G2U) synergy, where visual generation becomes an explicit intermediate reasoning step. Our framework enables a model to perform controlled generative acts, such as detail enhancement, context expansion or structural visualisation, to produce self-generated visual thoughts, which are then fed back into the model to refine perception without retraining or external tools. Through a comprehensive evaluation on twelve benchmarks, this reversed information flow consistently improves multimodal understanding. We show that generative fidelity bounds perceptual gain and that distinct families of edit prompts govern transfer efficiency. We further analyse whether models can decide what to imagine. While they can produce plausible edits, these self-generated visual thoughts lack stable task alignment, revealing that current large multimodal models fall short of true self-reflection. This work exposes a missing mechanism in unified cognition and suggests that imagination is not the end of understanding but its beginning.
arXiv:2605.15793v1 Announce Type: new Abstract: Pre-training neural operators on diverse partial differential equation (PDE) datasets has emerged as a promising direction for building general-purpose surrogate models in scientific machine learning. However, the inherent complexity and structural diversity of PDE solution operators make multi-PDE pre-training fundamentally challenging. Existing methods mainly address this by increasing model capacity, while leaving the target solution operators unchanged. Inspired by classical numerical analysis, we instead propose to transform complex and diverse solution operators into simpler, better-aligned forms that are easier to model jointly. Since the optimal transformation varies across PDE types, it must be adaptive and input-dependent, allowing a single neural operator to approximate an entire family of operators. We instantiate this idea as AOT-POT (adaptive operator-transformation for pre-training operator transformer), which expands hidden representations into multiple parallel streams, adaptively aggregates and redistributes them before and after each sub-layer, and mixes streams through Sinkhorn-projected doubly stochastic matrices for stable training. These mechanisms together reshape diverse solution operators into a unified form that can be effectively modeled by a single architecture. Empirically, AOT-POT achieves state-of-the-art performance on 12 PDE benchmarks with only 3\% additional parameters, reducing relative L2 error by up to 77.6\% (40.9\% on average). Fine-tuning AOT-POT further reduces L2 error by up to 92\% on in-domain PDEs and 89\% on out-of-domain PDEs (unseen types during pre-training), demonstrating that adaptive operator transformation is an effective and complementary direction for advancing PDE foundation models beyond simply scaling model capacity.
arXiv:2605.15794v1 Announce Type: new Abstract: We present ForMaT (Format-Preserving Multilingual Translation), a parallel corpus of 3,956 PDFs across 15 language pairs that preserves original layout metadata proposed for multimodal machine translation. To ensure structural diversity in the dataset, we employ K-Medoids sampling over 45 geometric features, capturing complex elements like nested tables and formulas to focus only on visually diverse PDF documents. Our evaluation reveals that current MT systems struggle with spatial grounding and geometric synchronization, often losing the link between text and its visual context. ForMaT provides a benchmark for developing layout-aware translation models that integrate visual and textual context for high-fidelity document reconstruction.
arXiv:2601.13529v2 Announce Type: replace Abstract: Cervical cancer accounts for a significant portion of the global cancer burden among women. Interstitial brachytherapy (ISBT) is a standard procedure for treating cervical cancer; it involves placing a radioactive source through a straight hollow needle within or in close proximity to the tumor and surrounding tissue. However, the use of straight needles limits surgical planning to a linear needle path. We present the OncoReach stylet, a handheld, tendon-driven steerable stylet designed for compatibility with standard ISBT 15- and 13-gauge needles. Building upon our prior work, we evaluated design parameters like needle gauge, spherical joint count and spherical joint placement, including an asymmetric disk design to identify a configuration that maximizes bending compliance while retaining axial stiffness. Free space experiments quantified tip deflection across configurations, and a two-tube Cosserat rod model accurately predicted the centerline shape of the needle for most trials. The best performing configuration was integrated into a reusable handheld prototype that enables manual actuation. A patient-derived, multi-composite phantom model of the uterus and pelvis was developed to conduct a pilot study of the OncoReach steerable stylet with one expert user. Results showed the ability to steer from less-invasive, medial entry points to reach the lateral-most targets, underscoring the significance of steerable stylets.
arXiv:2605.15798v1 Announce Type: new Abstract: Complex systems display emergent phenomena that vary significantly across spatial and temporal scales. These variations originate from fine-grained system processes, yet arriving at macroscopic dynamics from micro-level data -- particularly when large, high-resolution datasets are available -- remains a persistent challenge. Here we develop event-based spatiotemporal networks, a computational modelling framework that encodes system processes as discrete events anchored in space and time. Event-based spatiotemporal networks offer a unified, flexible and efficient approach to generate emergent behaviour in complex systems across space and time from these events. We demonstrate the effectiveness of event-based spatiotemporal networks through two illustrative real-world applications. First, following a local outbreak of a novel respiratory pathogen in the Netherlands, spatiotemporal networks enable fine-grained tracking of transmission routes and infection patterns through space and time. Second, we use spatiotemporal networks to model propagation of delays in a public transportation system (S-bahn) around Z\"urich, Switzerland. We also discuss broader uses of event-based spatiotemporal networks in fields like developmental biology and community ecology, where focusing on events rather than static system states can improve data analysis, simulation, and collection strategies.
arXiv:2511.08103v4 Announce Type: replace Abstract: Highly correlated photon sources can be realized through cooperative coupling among quantum systems, giving rise to superradiant collective emission. In solid-state ensembles, however, such collective behaviour is confined to subwavelength dimensions and is strongly suppressed at room temperature by inhomogeneous broadening and rapid dephasing, limiting practical implementations. Here, we show that molecular J-aggregates sustain room temperature superradiant emission and enter a highly collective regime when strongly coupled to delocalized photonic modes of a silicon bound-state-in-the-continuum (BIC) metasurface, extending J-exciton interactions far beyond the subwavelength limit. This enhanced cooperativity produces a photonic-fraction-dependent increase in emission rate and intensity and drives the system into a highly superbunched photon emission regime with g^((2))(0)>13. Spatial coherence measurements and stochastic modelling reveal that metasurface-mediated synchronization of ~10^3 J-excitons occurs within coupled superradiant domains spanning up to 6.7 um in diameter, corresponding to a 50-fold increase in inter-aggregate cooperative distance. These results establish exciton-polaritons in resonant dielectric metasurfaces as a platform to enhance superradiant emission and engineer temporally correlated light sources with picosecond-scale emission dynamics operating at room temperature.
arXiv:2511.04484v2 Announce Type: replace Abstract: We study the repeated optimal stopping problem, in which the same optimal stopping instance with an unknown distribution is solved repeatedly over $T$ rounds. We aim to simultaneously achieve strong per-round performance guarantees relative to a given baseline and sublinear regret across all rounds. Our primary contribution is a comprehensive theoretical characterization of whether and when these two objectives are compatible. First, under standard semi-bandit feedback, we prove that maintaining the per-round guarantee forces regret of $\Omega(T / \log T)$. Second, even under full feedback, we show that requiring almost-sure satisfaction of the per-round guarantee in every round is incompatible with sublinear regret. Third, under full feedback, we propose a general algorithmic framework that achieves both sublinear regret and the per-round guarantee with high probability. Our framework applies to canonical problems, including the prophet inequality, the secretary problem, and their variants under adversarial, random, and i.i.d. input models. For example, in the repeated prophet inequality problem, our method guarantees that, with high probability in each round, its expected reward is at least that of the classical single-sample algorithm, which achieves a $1/2$ competitive ratio, while simultaneously ensuring $\tilde{O}(\sqrt{T})$ regret. Furthermore, we establish a regret lower bound of $\Omega(\sqrt{T})$ even in the i.i.d. model, which is nearly tight with respect to the number of rounds.
Beyond the Training Data: Confidence-Guided Mixing of Parameterizations in a Hybrid AI-Climate Model
arXiv:2510.08107v5 Announce Type: replace Abstract: Persistent systematic errors in Earth system models (ESMs) arise from difficulties in representing the full diversity of subgrid, multiscale atmospheric convection and turbulence. Machine learning (ML) parameterizations trained on short high-resolution simulations show strong potential to reduce these errors. However, stable long-term atmospheric simulations with hybrid (physics + ML) ESMs remain difficult, as neural networks (NNs) trained offline often destabilize online runs. Training convection parameterizations directly on coarse-grained data is challenging, notably because scales cannot be cleanly separated. This issue is mitigated using data from superparameterized simulations, which provide clearer scale separation. Yet, transferring a parameterization from one ESM to another remains difficult due to distribution shifts that induce large inference errors. Here, we present a proof-of-concept where a ClimSim-trained, physics-informed NN convection parameterization is successfully transferred to ICON-A. The scheme is (a) trained on adjusted ClimSim data with subtracted radiative tendencies, and (b) integrated into ICON-A. The NN parameterization predicts its own error, enabling mixing with a conventional convection scheme when confidence is low, thus making the hybrid AI-physics model tunable with respect to observations and reanalysis through mixing parameters. This improves process understanding by constraining convective tendencies across column water vapor, lower-tropospheric stability, and geographical conditions, yielding interpretable regime behavior. In AMIP-style setups, several hybrid configurations outperform the default convection scheme (e.g., improved precipitation statistics). With additive input noise during training, both hybrid and pure-ML schemes lead to stable simulations and remain physically consistent for at least 20 years.
arXiv:2510.06062v2 Announce Type: replace Abstract: Reinforcement learning (RL) has shown great promise in large language models (LLMs) post-training, which typically rely on token-level clipping to maintain stability during optimization. Despite the empirical success of GRPO-style methods, we identify a fundamental and previously overlooked challenge in this popular Outcome-Supervised RL (OSRL) paradigm. We reveal that in OSRL, where advantages are shared across tokens within a response, importance sampling (IS) ratios deviate from their traditional purpose of distribution correction as in classic RL, which become token-level weights that allocate the shared advantage signal across tokens. We show that this hidden role shift induces a critical mismatch for positive-advantage tokens, leading to unbalanced token weighting between positive and negative tokens. Specifically, it suppresses the update of underrepresented tokens that are lagging behind, while over-amplifying already high-probability tokens. This mismatch results in rich-get-richer dynamics that over-reinforce confident tokens, weaken catch-up learning that drive entropy collapse, excessive repetition, and premature convergence. To address this, we propose Asymmetric Importance Sampling Policy Optimization (ASPO), a simple yet effective strategy that reverses the ratio-induced weighting of positive-advantage tokens, while stabilizing extreme updates and maintaining gradient flow. This mismatch correction aligns their update direction with the learning dynamics of negative ones. Comprehensive experiments across math reasoning and coding benchmarks demonstrate that ASPO significantly mitigates entropy collapse, improves training stability, and enhances performance over strong GRPO-based baselines. Our analysis provides new insights into the role of token-level weighting in OSRL and highlights the critical importance of correcting ratio-induced weighting in LLM RL.
arXiv:2605.15809v1 Announce Type: new Abstract: Symbolic regression (SR) aims to discover explicit mathematical expressions that explain observed data and is widely used in domains where interpretability is essential. Because interpretability requires expressions to reflect meaningful regularities, SR is sensitive to observations that deviate from the dominant relationship. Such irregular observations, or outliers, are common in real-world data and can hinder SR from identifying underlying regularities. Robust regression mitigates this by downweighting observations with large residuals. However, deciding which observations should be treated as outliers is often ambiguous and depends on user interpretation and domain knowledge, a perspective largely overlooked in existing SR studies. This motivates approaches that present multiple candidate expressions, allowing users to examine different residual patterns and choose expressions consistent with their expertise. We propose diversified residual symbolic regression (DRSR), which achieves high predictive accuracy while promoting diversity with respect to residual patterns based on the Quality-Diversity paradigm. DRSR collects multiple expressions that fit the data well but differ in how residuals are distributed, enabling post-search selection aligned with domain knowledge. On a synthetic mixture dataset, DRSR produces more diverse expressions than conventional SR while capturing multiple underlying relationships. On a real-world astronomical dataset, DRSR discovers multiple expressions consistent with known physical relationships.
arXiv:2605.15812v1 Announce Type: new Abstract: Recent advances in foundation models have enabled conversational agents that aim for sustained companionship rather than mere task completion. Yet most still remain unable to support natural, long-term companion-like interactions, resulting in experiences that feel episodic and inauthentic. We argue that current agents overlooked cross-temporal modeling of agents' social behaviors and internal emotions: generated behaviors rarely influence an agent's emotional state, and emotional states seldom shape subsequent behaviors. We present Cross-Temporal Emotion Modeling (CTEM), a framework that links long-term behavioral history to moment-to-moment emotional expression. CTEM establishes a closed loop where past experiences update an evolving emotional state; this state conditions immediate interactions; and user feedback continually revises both memory and emotional state, enabling reflection and anticipation. We instantiate CTEM as Auri, a companion agent on an instant-messaging platform, and report a 21-day in-the-wild study showing that CTEM shows improvements in perceived naturalness, coherence, and emotional harmony.
arXiv:2605.15226v1 Announce Type: new Abstract: We ask whether agentic AI systems built for software engineering transfer to realistic hardware engineering. Existing hardware LLM benchmarks isolate sub-tasks but none jointly requires repository navigation, hierarchy-aware localization, Electronic Design Automation (EDA) executable verification, and maintenance-style patching. We introduce \textbf{Phoenix-bench}, a synchronized corpus of 511 verified Verilator instances from 114 GitHub repositories, each shipped with the developer patch, design-flow labels, fail-to-pass and pass-to-pass testbenches, and a Docker-pinned EDA environment so resolved-rate differences reflect agent behavior rather than toolchain availability. Using Phoenix-bench we run a uniform evaluation of four commercial agents and eight open-source agentic structures across four LLM backbones, plus two diagnostic interventions (file-level oracle localization and one round of testbench-log feedback). Three findings emerge. (i)~Software and hardware are fundamentally different engineering tasks: the same agent loses 37\% to 58\% from SWE-bench Verified to Phoenix-bench because hardware bugs propagate across parallel instantiated modules through signal flow rather than along a software-style call graph, and software-tuned agents stop at the symptom file instead of tracing back through the instantiation chain. (ii)~Failures concentrate on design control-flow / finite state machine (FSM) bugs, verification testbench bugs, and hard cases that demand cross-hierarchy signal-flow tracking and coordinated multi-file edits. (iii)~Localization granularity matters far more than localization itself: a perfect file-level oracle yields only $+1.4$\% because the agent then breaks files that did not need editing, while a single round of test case feedback lifts resolved rate by $42$\% to $45$\% because the test case tells \emph{where} the bug is and \emph{what} the fix has to look like.
arXiv:2605.15224v1 Announce Type: new Abstract: Large language model-based agents make mistakes, yet critique can often guide the same model toward correct behavior. However, when critique is removed, the model may fail again on the same query, indicating that it has not internalized the critique's guidance into its underlying capability. Meanwhile, a frozen critic cannot improve its feedback quality over time, limiting the potential for iterative self-improvement. To address this, we propose learning to internalize self-critique with reinforcement learning(ICRL), a novel framework that jointly trains a solver and a critic from a shared backbone to convert critique-induced success into unassisted solver ability. The critic is rewarded based on the solver's subsequent performance gain, incentivizing actionable feedback. To address the distribution shift between critique-conditioned and critique-free behavior, ICRL introduces a distribution-calibration re-weighting ratio that selectively transfers critique-guided improvements compatible with the solver's own prompt distribution. Additionally, a role-wise group advantage estimation stabilizes joint optimization across the two roles. Together, these mechanisms ensure that the solver learns to improve itself without external critique, rather than becoming dependent on critique-conditioned behavior. We evaluate ICRL on diverse benchmarks spanning agentic and mathematical reasoning tasks, using Qwen3-4B and Qwen3-8B as backbones. Results show consistent improvements, with average gains of 6.4 points over GRPO on agentic tasks, and 7.0 points on mathematical reasoning. Notably, the learned 8B critic is comparable to 32B critics while using substantially fewer tokens. The code is available at https://github.com/brick-pid/ICRL.
arXiv:2405.11073v2 Announce Type: replace Abstract: We propose a semantic foundation for logics for reasoning in settings that possess a distinction between equality of variables, a coarser equivalence of variables, and a notion of conditional independence between variables. We show that such relations can be modelled naturally in atomic sheaf toposes. Equivalence of variables is modelled by a relation of atomic equivalence that is possessed by every atomic sheaf. We identify additional structure on the category generating the atomic topos that allows the relation of conditional independence to be interpreted in the topos. We then study the logic of equivalence and conditional independence that is induced by the internal logic of the topos. This atomic sheaf logic is a classical logic that validates a number of fundamental reasoning principles relating equivalence and conditional independence. As a concrete example of this abstract framework, we use the atomic topos over the category of surjections between finite nonempty sets as our main running example. In this category, the interpretations of equivalence and conditional independence coincide with those given by the multiteam semantics of independence logic, in which the role of equivalence is taken by the relation of mutual inclusion. A major difference from independence logic is that, in atomic sheaf logic, the multiteam semantics of the equivalence and conditional independence relations is embedded within a classical surrounding logic. We outline two other instances of our framework, to demonstrate its versatility. The first is a category of probability sheaves, in which atomic equivalence is equality-in-distribution, and the conditional independence relation is the usual probabilistic one. Our other example is the Schanuel topos where equivalence is orbit equality and conditional independence amounts to a relative form of separatedness.
arXiv:2505.07322v4 Announce Type: replace Abstract: High-Dynamic-Range Wide-Color-Gamut (HDR-WCG) technology is becoming increasingly widespread, driving a growing need for converting Standard Dynamic Range (SDR) content to HDR. Existing methods primarily rely on fixed tone mapping operators, which struggle to handle the diverse appearances and degradations commonly present in real-world SDR content. To address this limitation, we propose a generalized SDR-to-HDR framework that enhances robustness by learning attribute-disentangled representations. Central to our approach is Realistic Attribute-Disentangled Representation Learning (RealRep), which explicitly disentangles luminance and chrominance components to capture intrinsic content variations across different SDR distributions. Furthermore, we design a Luma-/Chroma-aware negative exemplar generation strategy that constructs degradation-sensitive contrastive pairs, effectively modeling tone discrepancies across SDR styles. Building on these attribute-level priors, we introduce the Degradation-Domain Aware Controlled Mapping Network (DDACMNet), a lightweight, two-stage framework that performs adaptive hierarchical mapping guided by a control-aware normalization mechanism. DDACMNet dynamically modulates the mapping process via degradation-conditioned features, enabling robust adaptation across diverse degradation domains. Extensive experiments demonstrate that RealRep consistently outperforms state-of-the-art methods in both generalization and perceptually faithful HDR color gamut reconstruction.
arXiv:2605.16008v1 Announce Type: new Abstract: Plaque assays remain the gold standard readout of virus infectivity; however, plaque counting from plate images is labor-intensive and prone to inter-operator variability. We present an end-to-end, computer-aided workflow for cytopathic effect-based virus titration directly from laboratory plaque assay images. The proposed approach combines two models derived from the Segment Anything Model (SAM): a SAM2-based well-segmentation module that localizes assay wells across heterogeneous imaging conditions, and a SAM-based plaque-segmentation model that detects and enumerates plaques within each well. The method was evaluated on a mixed dataset comprising private plaque assay images of Mayaro virus and Coxsackievirus B3, together with public Vaccinia virus images from the VACVPlaque dataset. The pipeline outputs per-well plaque counts, automatically computes plaque-forming units per milliliter (PFU/mL), and is integrated into a web-based platform that allows users to review results and organize experiments. On held-out plates (17 from MAYV/CVB3 and 22 from VACV), the workflow generalized across two plate formats (6-well and 12-well) and showed strong agreement with manual annotations (Pearson correlation coefficients of 0.92 for MAYV/CVB3 and 0.88 for VACV). Automated plaque counts were further compared with annotations from four independent experts, demonstrating high concordance. The proposed system will be open sourced and publicly released upon acceptance of this manuscript to enable reproducible, scalable, and audit-ready plaque assay analysis while substantially reducing manual annotation effort.