Academic Intelligence · Curated Daily

Explore the Frontier of Global Academia

AcademicHub aggregates real-time literature from top journals and preprint platforms. Build your personal research radar and let large language models compile cross-disciplinary analysis briefings automatically.

01.
arXiv (math.PR) 2026-06-16

Excursion Fluctuations and Spectral Universality in Gaussian Fields

arXiv:2606.15630v1 Announce Type: new Abstract: We study the large-scale spatial fluctuations of excursion volumes for a class of smooth stationary Gaussian fields. In the case of Berry's random wave model in dimension $d \geq 2$, we show that the spatial fluctuations for fixed $u>0$ converge to the fractional Gaussian field $(-\Delta)^{-1/4}W$ in the space of tempered distributions $\mathcal S'(\mathbb{R}^d)$, where $W$ is the $d$-dimensional Gaussian white noise. This explains the long-range correlations in the apparent filament structure of the Random Plane Wave model. For a class of smooth planar Gaussian fields whose spectral density has a power-law singularity at the origin, we prove convergence to fractional Gaussian fields with an index determined by the singularity exponent. More generally, the results illustrate that, for stationary random measures, large-scale spatial fluctuations are determined by the behaviour of the spectral measure density exponent near zero.

02.
Nature Biotechnology 2026-06-11

Large-scale, spatially resolved panoramic CRISPR screening in native tissue environments using Perturb-DBiT

Authors:

Spatially resolved CRISPR screening in vivo has been limited to small perturbation panels and subsets of protein-coding RNAs. We present Perturb-DBiT, a method for co-sequencing of spatial total RNA whole transcriptomes and single guide RNAs (sgRNAs) on the same tissue section in situ. In a human cancer metastatic colonization model, we applied large (80,000+) sgRNA panels across tumor colonies in multiple consecutive tissue sections alongside their corresponding total RNA transcriptomes. We linked perturbations affecting long noncoding RNA covariation, microRNA–mRNA interactions and distinct amino acid-specific tRNA alterations to tumor migration and growth. By integrating transcriptional pseudotime trajectories, we further observed the impact of perturbations on clonal dynamics and cooperation. In an immune-competent syngeneic mouse model, investigation of the tumor immune microenvironment indicated distinct, synergistic effects on immune infiltration and suppression. Perturb-DBiT provides a spatially resolved comprehensive view of perturbation responses in complex tissues, including small and large RNA regulation, tumor proliferation, migration, metastasis and immune interactions. In vivo CRISPR genetic perturbations are spatially mapped at scale.

03.
arXiv (CS.LG) 2026-06-12

Dense Supervision, Sparse Updates: On the Sparsity and Geometry of On-Policy Distillation

arXiv:2606.13657v1 Announce Type: new Abstract: On-policy distillation (\textsc{OPD}) has recently become a prominent post-training recipe as it combines two desirable ingredients: on-policy student trajectories and dense teacher supervision, yet how this hybrid changes a model's parameters remains unclear. Across several language and vision-language model pairs and use cases, our analysis yields two main findings. On sparsity, \textsc{OPD}-style updates are small and coordinate-sparse. They are distributed across layers and are usually FFN-heavy. This sparse structure is operationally useful: training only the discovered subnetwork recovers nearly the same performance as full \textsc{OPD}. However, the sparsity-inducing SGD optimizer underperforms AdamW in our optimizer ablation, likely because dense teacher supervision preserves heterogeneous coordinate-wise gradient scales where AdamW's adaptive scaling remains useful. On geometry, the updates are numerically full-rank but spectrally concentrated; they lie mostly away from the principal singular subspaces of the source weights and fall disproportionately on coordinates where the source weights are close to zero. These findings suggest that dense teacher supervision does not turn \textsc{OPD} into ordinary dense parameter rewriting; instead, \textsc{OPD} retains important geometric signatures of on-policy post-training.

04.
arXiv (quant-ph) 2026-06-15

An integrated ultrahigh vacuum cluster tool for diamond surface science and single nitrogen-vacancy center measurements

arXiv:2606.13961v1 Announce Type: new Abstract: We present a custom-designed ultrahigh vacuum (UHV) cluster tool developed for studying shallow nitrogen-vacancy (NV) centers in diamond, enabling in situ diamond surface preparation, characterization, and single NV center dynamics measurements within a single connected platform. The system combines a surface science chamber for controlled surface modification and analysis with a cryogenic confocal microscope chamber dedicated to NV spin and optical measurements. This integrated approach enables a direct correlation between diamond surface chemistry and the resulting NV spin and charge properties. The instrument provides a versatile platform for systematic studies of surface-induced decoherence mechanisms and charge dynamics for shallow NV centers, and establishes a pathway toward reproducible surface engineering for quantum sensing applications.

05.
arXiv (CS.AI) 2026-06-11

HERO: Hindsight-Enhanced Reflection from Environment Observations for Agentic Self-Distillation

arXiv:2606.11559v1 Announce Type: new Abstract: Reinforcement learning typically improves multi-turn agent capabilities through the terminal outcome of the trajectories, which makes it difficult to determine credit assignments for each intermediate turns. Recent on-policy self-distillation methods offer a promising alternative by converting privileged feedback into dense token-level supervision through a self-teacher. Our study is motivated by the unexpected performance degradation observed when naively extending this paradigm to multi-turn settings, which we attribute to a lack of alignment between privileged feedback, such as successful trajectories or terminal outcomes, and the student's current decision context. We introduce HERO, a hindsight-enhanced self-distillation framework that uses next environment observations as locally aligned feedback. After each rollout, HERO reflects on the completed interaction to convert each observation into a compact turn-level diagnosis, that captures actionable feedback about the original action such as its necessity, validity or failure cause. On TauBench and WebShop, HERO improves task success and reduces unnecessary turns over environment-feedback-only self-distillation and GRPO. It is especially effective under limited training turn budgets, where successful rollouts are rare and GRPO provides weak reward-contrast signals.

06.
arXiv (CS.LG) 2026-06-12

How Useful is Causal Invariance for Domain Adaptation in Finite-Sample Settings?

arXiv:2606.12680v1 Announce Type: new Abstract: Machine learning models often degrade when they are deployed on a target distribution that differs from the source distributions they were trained on. Recent work in causality-based domain generalization has shown how shared causal structure between domains can induce invariant predictors, e.g., models on a subset of features which have stable risk across structured domain shifts. However, the extent to which such population-level causal invariances can lead to gains in finite-sample settings remains underexplored. In particular, in practice we often have access to a few labeled target samples, a setting called supervised domain adaptation (sDA). In this paper, we explore when (full or partial) causal knowledge can provably improve supervised domain adaptation. As a first step, we study linear regression, where full or partial causal knowledge specifies a collection of invariant or possibly invariant feature subsets, each yielding a source-trained candidate predictor. We derive matching upper and lower bounds showing that finite-sample gains are governed by the target-risk margins separating the candidates, together with the finite-source estimation error. When these margins are sufficiently large relative to $n_Q$, an adaptive aggregation procedure can match the best candidate predictor while avoiding negative transfer relative to target-only learning. On the other hand, when the margins are too small, no algorithm can reliably exploit the candidate collection to obtain faster finite-sample rates. We further connect these margins to structural shift magnitude in linear SCMs and validate the theory on real-world causal benchmarks.

07.
arXiv (CS.LG) 2026-06-12

QoS Improvement in Multi User Cellular-Symbiotic Radio Network Assisted by Active-STAR-RIS

arXiv:2401.08301v2 Announce Type: replace-cross Abstract: In this article, we employ active simultaneously transmitting and reflecting reconfigurable intelligent surfaces (ASRIS) to enhance the quality of 6G cellular network services. The network integrates commensal symbiotic radio (CSR) subsystems to facilitate communication between passive Internet of Things (IoT) users and active users, referred to as symbiotic backscatter devices (SBDs) and symbiotic user equipments (SUEs), respectively. Since the SBDs are passive, transmitting information to the SUEs poses significant challenges. To overcome this challenge, we harness the capabilities of massive multiple input multiple output (MIMO) antennas within the base station (BS) to relay the information transmitted by SBDs with greater power. This scheme uses the non-orthogonal multiple access (NOMA) technique for multiple access among all users, and potential interferences are eliminated using successive interference cancellation (SIC). The primary objective is to maximize the throughput between SBDs and SUEs. To achieve this, we formulate an optimization problem involving variables such as active beamforming coefficients at the BS and ASRIS, phase adjustments of ASRIS, and scheduling parameters between CSR and cellular networks. To solve this optimization problem, we used three deep reinforcement learning (DRL) methods: proximal policy optimization (PPO), twin delayed deep deterministic policy gradient (TD3), and asynchronous advantage actor critic (A3C). These methods were simulated, and the results demonstrate that A3C, TD3, and PPO have the best convergence speeds and achieve the highest increases in network throughput, respectively. Finally, the proposed scheme was evaluated using passive simultaneously transmitting and reflecting RIS (STAR-RIS), which demonstrated poorer performance compared to ASRIS.

08.
arXiv (quant-ph) 2026-06-15

A new class of degenerate solutions to the massless Dirac equation and their potential applications in optical memories

arXiv:2606.14256v1 Announce Type: new Abstract: In this article, we present a novel class of degenerate solutions to the massless Dirac equation, corresponding to a wide variety of electromagnetic 4-potentials and fields, including both zero field and circularly polarized electromagnetic waves. An interesting property of these solutions is that the spin of the particles rotates in synchronization with the electric and magnetic fields of the electromagnetic waves. These results could be utilized for the development of optical memories based on materials supporting massless Dirac fermions, such as graphene.

09.
arXiv (CS.CL) 2026-06-16

REFLEX: Reflective Evolution from LLM Experience

Authors:

Large multimodal language models (LLMs) have emerged as powerful tools for guiding evolutionary search toward interpretable programmatic policies. However, existing frameworks rely on a monolithic model call to simultaneously interpret visual behavioral evidence and synthesize corrective code. This diagnosis-repair entanglement creates an opaque feedback loop, obscuring the rationale behind mutations and preventing the retention of algorithmic insights across independent runs. To achieve auditable and efficient policy search, we argue that visual diagnosis must be structurally decoupled from code generation. We present REFLEX, a train-free evolutionary framework that operationalizes this decoupling. In REFLEX, a vision-enabled Critic first distills task-specific behavioral evidence into structured, auditable diagnoses. Subsequently, a text-optimized Actor synthesizes child policies using these diagnoses alongside a persistent, self-evolving Skill Memory of reusable code snippets. This architecture not only provides transparent mutation traces but also enables cross-run programmatic knowledge transfer. Extensive evaluations across control benchmarks (Lunar Lander, Acrobot, Pendulum) and a 36-dimensional antenna array synthesis task demonstrate exceptional sample efficiency. Notably, REFLEX solves Acrobot and Pendulum in under 10 LLM calls and reaches a best Normalized Weighted Score of 1.092 on Lunar Lander, achieving highly competitive final performance while significantly accelerating the early-stage discovery of transparent policies.

10.
arXiv (CS.AI) 2026-06-18

Generating Natural and Expressive Robot Gestures through Iterative Reinforcement Learning with Human Feedback using LLMs

arXiv:2606.18747v1 Announce Type: cross Abstract: Expressive gestures are essential for natural and effective communication, complementing speech when verbal cues alone are insufficient (e.g., pointing). For social robots such as the humanoid Pepper, producing natural and expressive movements is critical for improving human-robot interaction (HRI) and long-term acceptance. However, generating gestures remains challenging due to reliance on expert-authored animations, resulting in rigid behaviors that are impractical for dynamic and diverse environments. Alternatively, machine learning approaches often struggle to capture perceived naturalness, becoming increasingly challenging with more degrees of freedom. Consequently, producing expressive robot gestures requires a system that can adapt to the environment while adhering to social norms and physical constraints. Recent advances in large language models (LLMs) enable dynamic code generation, offering new opportunities for runtime gesture synthesis from natural language. In this paper, we integrate ChatGPT into the humanoid robot Pepper to generate co-speech gestures aligned with conversational output. While this baseline enables flexible gesture generation, the resulting motions are often perceived as stiff and unnatural. To address this limitation, we introduce an iterative reinforcement learning with human feedback (RLHF) system that finetunes gesture generation based on user evaluations, leveraging an iterative user study to compare Pepper's generated gestures. Our results show that RLHF improved the LLM's co-speech generative capabilities, producing more expressive, relevant and fluid movements.

12.
arXiv (CS.CV) 2026-06-16

UniT: Unified Multimodal Chain-of-Thought Test-time Scaling

Unified models can handle both multimodal understanding and generation within a single architecture, yet they typically operate in a single pass without iteratively refining their outputs. Many multimodal tasks, especially those involving complex spatial compositions, multiple interacting objects, or evolving instructions, require decomposing instructions, verifying intermediate results, and making iterative corrections. While test-time scaling (TTS) has demonstrated that allocating additional inference compute for iterative reasoning substantially improves language model performance, extending this paradigm to unified multimodal models remains an open challenge. We introduce UniT, a framework for multimodal chain-of-thought test-time scaling that enables a single unified model to reason, verify, and refine across multiple rounds. UniT combines agentic data synthesis, unified model training, and flexible test-time inference to elicit cognitive behaviors including verification, subgoal decomposition, and content memory. Our key findings are: (1) unified models trained on short reasoning trajectories generalize to longer inference chains at test time; (2) sequential chain-of-thought reasoning provides a more scalable and compute-efficient TTS strategy than parallel sampling; (3) training on generation and editing trajectories improves out-of-distribution visual reasoning. These results establish multimodal test-time scaling as an effective paradigm for advancing both generation and understanding in unified models.

13.
arXiv (CS.AI) 2026-06-19

LLM Doesn't Know What It Doesn't Know: Detecting Epistemic Blind Spots via Cross-Model Attribution Divergence on Clinical Tabular Data

arXiv:2606.19509v1 Announce Type: new Abstract: Large language models (LLMs) are increasingly applied to structured clinical data, yet whether they can recognize the limits of their own knowledge on such tasks remains unexplored. We study this question through the lens of cross-model attribution divergence with the goal of reducing epistemic uncertainty for structured tasks, comparing Qwen 2.5 7B and XGBoost on a prediction task via attribution divergence analysis. We report four findings. First, LLM verbalized confidence is epistemically vacuous, it outputs a near-constant (0.856-0.937) regardless of whether accuracy is 49% or 75.3%, tracking prompt format rather than prediction quality. Second, the LLM exhibits an inverse difficulty effect: accuracy drops to 64.8% when XGBoost is 99% correct, but matches XGBoost (73.8% vs. 73.1%) when it is moderately uncertain. Third, few-shot examples and SHAP-derived feature evidence are orthogonal, super-additive interventions: they reduce the Attribution Disagreement Score (ADS) from 1.54 to 0.38 and improve accuracy from 49% to 75.3% without training. Fourth, a cross-model calibrator that determined LLM reliability using attribution divergence signals reduces expected calibration error from 0.254 to 0.080, replacing uninformative verbalized confidence with patient-specific reliability estimates, without accessing model internals or requiring repeated inference. We frame these findings as a cold start problem for LLMs on structured data and outline a path toward genuine epistemic self-awareness.

14.
arXiv (CS.AI) 2026-06-18

A Taxonomy of Mental Health and Technology Needs for Alzheimer's and Dementia Caregivers

arXiv:2606.19247v1 Announce Type: cross Abstract: Family members caring for individuals with Alzheimer's disease and related dementias (AD/ADRD) provide the foundation of long-term care worldwide. In 2023, more than 11 million U.S. family and friends contributed 18 billion hours of unpaid care, often at the cost of their own physical and mental health. These informal caregivers – also referred as the "invisible second patients" – experience elevated rates of mental health problems. Yet research commonly reduces their complex psychosocial experiences to a single construct of caregiver burden, obscuring which specific needs are unmet or effectively supported. At the same time, digital and AI-enabled technologies are rapidly expanding, from smartphone apps and videoconferencing to sensor platforms and AI chatbots. However, the absence of shared frameworks across medicine, psychology, and technology research limits cumulative progress. This study introduces a Caregiver Mental Health and Technology Taxonomy that systematically links AD/ADRD caregiver needs with corresponding classes of technology-based interventions. Drawing from an interdisciplinary literature review and two qualitative studies with caregivers, the taxonomy identifies mismatches between caregiver priorities and existing technological support, highlights under-served domains such as relational strain and compassion fatigue, and proposes design directions for adaptive, responsive systems. The framework offers a shared vocabulary to guide clinicians, researchers, and technology designers in developing more person-centered and clinically grounded innovation in dementia care.

15.
arXiv (CS.AI) 2026-06-12

LLMs as ASP Programmers: Self-Correction Enables Task-Agnostic Nonmonotonic Reasoning

arXiv:2604.27960v2 Announce Type: replace Abstract: Recent large language models (LLMs) have achieved impressive reasoning milestones but continue to struggle with high computational costs, logical inconsistencies, and sharp performance degradation on high-complexity problems. While neuro-symbolic methods attempt to mitigate these issues by coupling LLMs with symbolic reasoners, existing approaches typically rely on monotonic logics (e.g., SMT) that cannot represent defeasible reasoning – essential components of human cognition. We present "LLM+ASP," a framework that translates natural language into Answer Set Programming (ASP), a nonmonotonic formalism based on stable model semantics. Unlike prior "LLM+ASP" approaches that require manually authored knowledge modules, domain-specific prompts, or evaluation restricted to single problem classes, our framework operates without any per-task engineering and applies uniformly across diverse reasoning tasks. Our system utilizes an automated self-correction loop where structured feedback from the ASP solver enables iterative refinement. Evaluating across six diverse benchmarks, we demonstrate that: (1) stable model semantics allow LLMs to naturally express default rules and exceptions, outperforming SMT-based alternatives by significant margins on nonmonotonic tasks; (2) iterative self-correction is the primary driver of performance, effectively replacing the need for handcrafted domain knowledge; (3) compact in-context reference guides substantially outperform verbose documentation, revealing a "context rot" phenomenon where excessive context hinders constraint adherence.

16.
arXiv (CS.AI) 2026-06-16

Sensory Restoration via Brain-Computer Interfaces: A Unified 2 x 2 Framework and Convergence Roadmap

Authors:

arXiv:2606.15091v1 Announce Type: cross Abstract: Millions of individuals worldwide suffer from sensory and communication deficits caused by neurodegenerative diseases, stroke, or trauma. Brain-computer interfaces (BCIs) offer a promising avenue for sensory and motor restoration. However, the scientific literature remains highly fragmented between invasive neuroprosthetics and non-invasive electrophysiological decoders, with a lack of consistent terminology and comparison metrics. This chapter proposes a unified 2 x 2 framework categorizing BCIs along two axes: degree of invasiveness (invasive vs. non-invasive) and signal direction (afferent sensory-IN vs. efferent sensory-OUT). We define and distinguish the paradigms of restoration, substitution, and augmentation. Furthermore, we outline a structural roadmap for the convergence of these modalities over near-, medium-, and long-term horizons, focusing on physical limits and the integrative role of machine learning foundation models.

17.
arXiv (CS.AI) 2026-06-12

PhononBench:A Large-Scale Phonon-Based Benchmark for Dynamical Stability in Crystal Generation

arXiv:2512.21227v3 Announce Type: replace-cross Abstract: In recent years, generative artificial intelligence has made significant advances in the design of crystalline materials, giving rise to approaches based on graph neural networks, diffusion models, and large language models. Existing evaluations commonly follow the stability-uniqueness-novelty (S.U.N.) framework, where stability is primarily assessed using thermodynamic criteria, which do not fully capture the dynamical stability essential for a material's practical existence. Dynamical stability is a key determinant of whether a material can be synthesized and persist, with phonon spectrum calculations serving as the standard for its evaluation. However, the high computational cost of such calculations has prevented large-scale assessment of dynamical stability in generated crystals. In this work, we introduce PhononBench, the first large-scale benchmark for dynamical stability in AI-generated crystals. Leveraging the recently developed MatterSim interatomic potential, which achieves density-functional-theory (DFT)-level accuracy in phonon predictions across more than 10,000 materials, PhononBench enables efficient phonon calculations and dynamical-stability analysis for 133,838 crystal structures generated by 7 leading crystal generation models. PhononBench reveals a widespread limitation of current generative models: unless otherwise specified, all reported dynamical-stability metrics are evaluated at a phonon-frequency threshold of -0.1 THz, with the average dynamical-stability rate across all generated structures being only 32.15%, and the top-performing model, MatterGen, reaching just 45.05%.In addition, we identify 32,995 crystal structures that are phonon-stable across the entire Brillouin zone under a strict threshold of -0.001 THz. In addition, a web-based service is accessible at http://phononbench.cn/, enabling minute-level ultra-fast phonon predictions.

18.
arXiv (CS.CL) 2026-06-16

PVminerLLM2: Improving Structured Extraction of Patient Voice via Preference Optimization

Motivation: Patient-generated text contains critical information on patients' lived experiences, social context, and care engagement, but remains largely unstructured, limiting its use in patient-centered outcomes research. Prior work introduced the PV-Miner benchmark and PVMinerLLM models for structured extraction. However, supervised fine-tuning (SFT) alone struggles with rare, fine-grained, and unevenly distributed errors, particularly in token-critical structured outputs. Results: We present PVminerLLM2, an improved set of LLMs for structured patient voice extraction that applies preference optimization to address token-critical errors beyond the reach of supervised fine-tuning. Our method introduces (i) a preference objective with token-level gated stabilization term that prevents degradation of absolute token likelihood under preference optimization, and (ii) confusion-aware preference pair construction to better capture low-separation distinctions. We further incorporate token-importance weighting and inverse-frequency reweighing to address token imbalance and class skew. Across multiple model sizes, PVMinerLLM2 consistently outperforms strong baselines, achieving gains of up to 4.43% (Code), 3.50% (Sub-code), and 1.55% (Span), and outperforms baseline LLM trained with existing preference optimization methods. Availability and Implementation: The supplementary material, code, evaluation scripts, and trained models for PVminerLLM2 are publicly available at: https://github.com/Data-Mining-Lab-Yale/PVminerLLM2

19.
arXiv (CS.AI) 2026-06-12

Algorithmic Constitutionalism

arXiv:2606.12437v1 Announce Type: cross Abstract: The increasing encroachment of artificial intelligence (AI) on social life raises significant risks for society, particularly within the infospheres created and controlled by companies such as Google, Facebook, Apple, and Amazon. This article examines these risks through an in-depth analysis of Facebook's content moderation regime, which is already partially governed by algorithms. We argue that the idea of ethical engineering, often proposed in the literature as a solution to the governance challenges posed by AI, is inadequate for several reasons. In response, we develop an alternative framework, which we term "algorithmic constitutionalism." Our approach rests on three pillars: (a) a layered architecture consisting of two levels of code: (i) an operative or object level and (ii) a meta level designed to protect the system's core principles from algorithmically initiated change; (b) algorithmic meta-reasoning, which enables the system to operate simultaneously at both levels so that it can monitor, verify, and potentially correct in real time operations at the object level that depart from principles protected at the meta-code level; and (c) correction through deliberation. The article elaborates the concept of algorithmic constitutionalism and demonstrates how it may be applied to Facebook's content moderation regime. As part of this analysis, we examine the tension between societal constitutionalism and algorithmic constitutionalism. Paradoxically, attempts to subject AI systems to external deliberative control may also enable AI agents to intervene in that process, potentially undermining its purpose. The article concludes by considering the implications of this argument for the European Digital Services Act, which entered into force in October 2022.

21.
arXiv (CS.LG) 2026-06-12

The Stable Recovery Manifold: Geometric Principles Governing Recoverability in Continual Learning

arXiv:2606.13637v1 Announce Type: new Abstract: Catastrophic forgetting is often viewed as the destruction of previously learned knowledge during sequential learning. Building on the Accessibility Collapse framework, we investigate the geometric structure of recoverability in continual learning. Using Split CIFAR-100 and a sequentially trained ResNet-18, we analyze recoverability, representational drift, and recovery complexity across ten tasks. We introduce Recovery Subspace Dimensionality (k_t), a measure of the minimum number of singular directions required to preserve 90 percent of full probe performance. Contrary to our Recoverability Diffusion hypothesis, recovery dimensionality remains stable throughout training (mean k_t = 8.0) despite substantial representational drift. Principal-angle drift strongly predicts recoverability (r = -0.862), and a simple geometric model explains 82.2 percent of recoverability variance. These findings support the Stable Recovery Manifold hypothesis, suggesting that forgotten knowledge remains compactly decodable despite representational reorganization. The results indicate that catastrophic forgetting is primarily an accessibility and manifold-alignment problem rather than information destruction.

22.
arXiv (quant-ph) 2026-06-12

From 2D Yang-Mills to Calogero-Sutherland via a colored particle

arXiv:2606.13388v1 Announce Type: cross Abstract: We study Yang-Mills theory coupled to a particle on a cylinder, where gauge invariance and compactness reduce the dynamics to a finite dimensional quantum system. In the Abelian case, this yields a model equivalent to the Landau problem on a torus, with a degenerate ground state structure. We generalize this construction to non-Abelian gauge groups and show that, for SU(N), the system reduces to a one dimensional quantum many body problem with a singular Calogero-Sutherland-type interaction.

23.
arXiv (CS.CL) 2026-06-12

Entropy-Gradient Inversion: Moving Toward Internal Mechanism of Large Reasoning Models

The advancement of Large Reasoning Models (LRMs) has catalyzed a paradigm shift from reactive ``fast thinking'' text generation to systematic, step-by-step ``slow thinking'' reasoning, unlocking state-of-the-art performance in complex mathematical and logical tasks. However, the field faces the fundamental gap between token-level behavioral analysis and internal reasoning mechanisms, and the instability of reinforcement learning (RL) for reasoning optimization relying on costly external verifiers. We identify and formally define Entropy-Gradient Inversion, a robust negative correlation between token entropy and logit gradients that acts as a definitive geometric fingerprint for LRM reasoning capability. Building on this, we propose Correlation-Regularized Group Policy Optimization (CorR-PO), which embeds this inversion signature into RL reward regularization. Extensive experiments on various reasoning benchmarks across multiple model scales show CorR-PO consistently outperforms state-of-the-art baselines, confirming that stronger inversion directly correlates with superior reasoning performance.

24.
arXiv (CS.LG) 2026-06-11

AsFT: Anchoring Safety During LLM Fine-Tuning Within Narrow Safety Basin

arXiv:2506.08473v4 Announce Type: replace Abstract: Fine-tuning large language models (LLMs) improves performance but introduces critical safety vulnerabilities: even minimal harmful data can severely compromise safety measures. We observe that perturbations orthogonal to the alignment direction - defined by weight differences between aligned (safe) and unaligned models - rapidly compromise model safety. In contrast, updates along the alignment direction largely preserve it, revealing the parameter space as a "narrow safety basin". To address this, we propose AsFT (Anchoring Safety in Fine-Tuning) to maintain safety by explicitly constraining update directions during fine-tuning. By penalizing updates orthogonal to the alignment direction, AsFT effectively constrains the model within the "narrow safety basin," thus preserving its inherent safety. Extensive experiments on multiple datasets and models show that AsFT reduces harmful behaviors by up to 7.60%, improves task performance by 3.44%, and consistently outperforms existing methods across multiple tasks.

25.
arXiv (CS.CL) 2026-06-19

Characterizing Narrative Content in Web-scale LLM Pretraining Data

The narrative composition of web-scale LLM pretraining corpora remains largely unexplored even though narrative is a fundamental mode of human communication. We present the first fine-grained study of narrative features in Dolma, a 3-trillion-token open pretraining corpus. Drawing on narrative theory, we design a framework spanning three core narrative elements (agency, setting, and events) operationalized as 11 interpretable dimensions. After sampling and annotating a diverse set of 400 passages, we finetune and validate NarraBERT, a RoBERTa-based model for fine-grained narrative prediction. We apply NarraBERT to 3M passages, resulting in a new dataset, NarraDolma. We find (i) narrative structure is measurable at scale across extremely heterogeneous data, (ii) we uncover a continuous, multidimensional narrative structure underlying web text, and (iii) narrative qualities are unequally distributed across pretraining sources and topics in ways that current curation practices neither measure nor account for. Our framework, dataset, and analyses provide a foundation for understanding how narrative qualities are distributed in LLM pretraining data and for studying how data composition affects narrative reasoning tasks. We publicly release NarraDolma and NarraBERT.