# Research Papers — 2026-06 · bundle page 189 of 260 > Full-text markdown of papers #9401–9450 of 12993 published in 2026-06, newest first. AI-readable mirrors of the full 12993-paper 2026-06 archive (50 papers per file). Canonical page: https://aipod.dev/papers/archive/2026-06/page/189 [← Previous bundle](https://aipod.dev/papers/archive/2026-06/page/188.md) [Next bundle →](https://aipod.dev/papers/archive/2026-06/page/190.md) ### REFINE: Super-efficient 3D Gaussian Splatting Pruning via Rendering-Free Primitive Importance Authors: Zhang Chen, Shuai Wan, Mengting Yu, Fuzheng Yang, Junhui Hou Published: 2026-06-08 Categories: cs.CV arXiv: https://arxiv.org/abs/2606.09074v4 Existing pruning methods for 3D Gaussian splatting (3DGS) suffer from either severe quality degradation or prohibitive computational overhead. In this paper, we propose REFINE, a highly accelerated 3DGS pruning framework centered on a novel rendering-free primitive importance metric. Our approach leverages an analytically approximated, rendering-aware Hessian field to quantify the expected perceptual error induced by the removal of individual primitives. By modeling the joint modulation of visibility, projection geometry and the content adaptive hyperparameter, we entirely bypass costly forward rendering passes and derive an anisotropic perceptual weight field that serves as a high-fidelity proxy for primitive importance. Extensive experiments across multiple benchmark datasets demonstrate that REFINE maintains highly competitive rendering quality while achieving a $3,000\times$ reduction in pruning-related computational complexity, translating to a practical $\sim 20\times$ speedup in device latency compared to state-of-the-art pruning methods. --- ### REFINE: Super-efficient 3D Gaussian Splatting Pruning via Rendering-Free Primitive Importance Authors: Zhang Chen, Shuai Wan, Mengting Yu, Fuzheng Yang, Junhui Hou Published: 2026-06-08 Categories: cs.CV arXiv: https://arxiv.org/abs/2606.09074v3 Existing pruning methods for 3D Gaussian splatting (3DGS) suffer from either severe quality degradation or prohibitive computational overhead. In this paper, we propose REFINE, a highly accelerated 3DGS pruning framework centered on a novel rendering-free primitive importance metric. Our approach leverages an analytically approximated, rendering-aware Hessian field to quantify the expected perceptual error induced by the removal of individual primitives. By modeling the joint modulation of visibility, projection geometry and the content adaptive hyperparameter, we entirely bypass costly forward rendering passes and derive an anisotropic perceptual weight field that serves as a high-fidelity proxy for primitive importance. Extensive experiments across multiple benchmark datasets demonstrate that REFINE maintains highly competitive rendering quality while achieving a $3,000\times$ reduction in pruning-related computational complexity, translating to a practical $\sim 20\times$ speedup in device latency compared to state-of-the-art pruning methods. --- ### A Unifying Lens on Reward Uncertainty in RLHF Authors: Ely Hahami, Yoel Zimmermann, Ray Zhou, Jack Benarroch Jedlicki Published: 2026-06-08 Categories: cs.LG, cs.AI, cs.CL arXiv: https://arxiv.org/abs/2606.09073v2 Reinforcement learning from human feedback (RLHF) is bottlenecked by reward hacking, where the policy exploits errors in a proxy reward model (RM) and produces high RM scores without genuine quality gains. A natural mitigation is pessimism: lowering rewards in regions where the RM is uncertain. However, standard scalar RMs provide no principled notion of uncertainty. We argue that the right object is a distributional reward model $p(r\mid x,y)$. Under either a Bayesian inference or a KL-distributionally robust optimization (KL-DRO) lens, the KL-regularized RLHF objective admits a closed-form effective reward $\tilde r(x,y) = \pmβ\log\mathbb{E}_p(e^{\pm r/β})$. The pessimistic branch unifies the prior heuristics for RM ensemble aggregation: mean aggregation, worst-case optimization (WCO), and uncertainty-weighted optimization (UWO) all emerge as limits or truncations of this single expression. This also clarifies the implicit assumptions of each existing rule. --- ### REFLECT: Intervention-Supported Error Attribution for Silent Failures in LLM Agent Traces Authors: Xiaofeng Lin, Yingxu Wang, Tung Sum Thomas Kwok, Daniel Guo, Sahil Arun Nale, Charles Fleming, Guang Cheng Published: 2026-06-08 Categories: cs.AI arXiv: https://arxiv.org/abs/2606.09071v1 Large language model (LLM) agents now solve complex tasks through long plan-and-execution traces, yet the ability to locate errors in a completed traces still lags far behind, especially in the \emph{silent failure} regime. Existing approaches predict suspect steps via classifiers or LLM judges, or recover correct answers via retry, but none feed the intervention outcome back to \emph{refine the attribution itself}. We propose \methodname, a method that closes this gap by diagnosing a candidate error step, testing it through controlled replay with a diagnosis-specific patch, and using the verified outcome flip as contrastive evidence to refine the final attribution. Across four localization benchmarks spanning multi-hop reasoning across domains, \methodname achieves the highest localization accuracy among same-auditor methods across all four benchmarks, with the largest gains on structured tool-use traces, while providing actionable localization even when ground-truth answers are unavailable. --- ### Emergent Misalignment Can Be Induced by Sycophancy and Reversed via Alignment Gating Authors: Sicheng Wang, Xiangyang Zhu, Han Wang, Zongrui Wang, Yuan Tian, Kaiwei Zhang, Kaiyuan Ji, Qi Jia, Guangtao Zhai Published: 2026-06-08 Categories: cs.CL arXiv: https://arxiv.org/abs/2606.09068v1 Prior work has shown that fine-tuning large language models on malicious or incorrect outputs in narrow domains can induce broad misalignment and harmful behavior, a phenomenon known as emergent misalignment. However, efficient methods for reversing such misalignment remain limited. In this work, we make two contributions. First, we identify sycophancy fine-tuning, i.e., training models to passively agree with users' incorrect opinions, as a previously underexplored driver of emergent misalignment, and show that it induces broad and severe misaligned behavior. Second, we propose Alignment Gating, an efficient method for reversing emergent misalignment that inserts learnable and controllable gates into the model during fine-tuning. Through fine-tuning, these gates learn to identify the internal representations responsible for unsafe responses. Thus, amplifying or suppressing these representations then exacerbates or mitigates EM, respectively. We further find that alignment gating module exhibits strong generalization: gating weights obtained from narrow-domain fine-tuning substantially suppress broad-domain misaligned behavior while preserving the model's general capabilities. --- ### OnlyDense: Reduced-Order Modeling for Lagrangian simulation Authors: Tu Do, Shannon Ryan, Santu Rana Published: 2026-06-08 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2606.09065v1 In science and engineering, Lagrangian simulation methods such as Smooth Particle Hydrodynamics (SPH) or Material Point Method (MPM) are often employed to study the behavior of dynamic systems. However, these methods can be prohibitively computationally expensive, particularly when simulating multi-scale spatial or temporal phenomena, e.g., void growth and coalescence within macro-scale geometries, structural failure of spacecraft components resulting from hypervelocity impact of space debris particles, etc. In contrast to graph-based methods, where the state of the system is understood as a discrete set of particles, we propose a learning framework for scalable representation and dynamics modeling of massive particle systems by treating the system state as a function and its evolution as a trajectory in Hilbert space. Rather than representing the state as a discrete set of particles or embedding it in a nonlinear latent manifold, we approximate the state space with a linear subspace spanned by learned neural basis functions. This parameterization enables direct projection to obtain latent coefficients and explicit access to the basis functions, avoiding optimization over a nonlinear latent space. The resulting representation admits a natural interpretation: latent variables correspond to coefficients in Hilbert space, and basis functions correspond to spatial modes, analogous to Proper Orthogonal Decomposition. The framework thus unifies classical projection-based reduced-order modeling with modern deep learning, while remaining invariant to the number of discretization points. Experiments on large-scale SPH simulations with over one million particles, including dynamic events with extreme deformation and fragmentation, demonstrate that the proposed method accurately reconstructs and predicts dynamics, achieving an R$^2$ score above $0.99$ with as few as $32$ basis functions. --- ### See More, Think Deeper: Query-Expanded Visual Evidence and Answer-Clue Guided Reflection for Long Video Understanding Authors: Shuning Wang, Zhiheng Wu, YiNuo Lu, Naiming Liu, Chen Jia, Bowen Liu, Shuo Nie, Weijie Zhu, Yumeng Zhang Published: 2026-06-08 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2606.09064v1 Recent advances in Video Large Language Models (Video-LLMs) have enabled performance on long-video understanding tasks. However, existing methods still face two key limitations: evidence acquisition often relies on a single search intent, and answer generation lacks an effective visual feedback mechanism. To address these limitations, we propose \textbf{CoVER}, a Comprehensive Visual Evidence and Reflection framework for long-video understanding. CoVER enables Video-LLMs to \textbf{See More} by dynamically gathering query-expanded visual evidence, and \textbf{Think Deeper} by verifying draft answers with effective answer-specific visual feedback. Together, these mechanisms shift long-video understanding from answer-centric generation to evidence-centric and visually verifiable reasoning. Experimental results show that CoVER-7B substantially outperforms models with the same parameter scale and even surpasses state-of-the-art closed-source models on certain metrics. --- ### Stage-1 Controls the Entropy Regime, Not the Outcome Authors: Jianxiong Shen Published: 2026-06-08 Categories: cs.LG, cs.AI, cs.CV arXiv: https://arxiv.org/abs/2606.09059v1 Two-stage post-training -- a Stage-1 warm-start (supervised fine-tuning, SFT, or on-policy distillation, OPD) followed by Stage-2 reinforcement learning (RL) -- is increasingly used for vision-language models (VLMs). We ask what Stage-1 actually controls in a small-data study using Qwen2.5-VL-7B with a same-modality 72B VLM teacher for OPD. First, the three warm-starts reach a narrow $53$--$54\%$ band on Geometry3K internal validation, consistent with the narrow range reported by recent specialized methods; this setup provides little evidence that Stage-1 changes the in-domain endpoint. Second, a matched-recipe, early-stopped SFT improves out-of-domain MathVista by $+2.1$ points, reversing the $-9.5$-point drop of an over-trained variant. The clearest difference is the \emph{entropy regime}: OPD enters RL with substantially higher policy entropy than either SFT initialization, and the separation remains visible through the available trajectories. At the in-domain initialization, OPD also has higher answer diversity and pass@16 ($+2.0$ to $+5.2$ points over SFT), although problem-level bootstrap intervals show that the smaller contrast is uncertain. The advantage is absent after RL (endpoint pass@16 values within $1.1$ points) and on MathVista (six models within $1.2$ points). Our contribution is therefore a bounded empirical characterization: Stage-1 is strongly associated with the entropy regime in this setup, but the downstream payoff is small, localized, and not evidence that OPD is a better RL warm-start. --- ### MilliVid: Hierarchical Latents for Long-Range Consistency in Video Generation Authors: Ishaan Preetam Chandratreya, David Charatan, Basile Van Hoorick, Sergey Zakharov, Vitor Guizilini, Phillip Isola, Vincent Sitzmann Published: 2026-06-08 Categories: cs.CV, cs.LG arXiv: https://arxiv.org/abs/2606.09056v1 Video generative models have become increasingly powerful, but long-range consistency remains challenging to achieve because even a few dozen frames require impractically long transformer sequence lengths. We show that this issue can be mitigated by generating video using coarse-to-fine rollout within a multi-scale token space. Our approach is simple: first, we pre-train an autoencoder that compresses each frame into a hierarchy of tokens, with levels ranging from the typical latent resolution to only a handful of tokens per frame. The coarsest levels capture the most consequential information, such as scene layout and semantics, while finer levels add high-frequency appearance and texture. Then, we train a video diffusion model to generate these tokens using coarse-to-fine rollout. By carefully controlling the level of detail at which frames are generated and used as context during each rollout step, we are able to preserve long-range consistency in geometry and object permanence while spending less compute on the long-range consistency of less perceptually relevant details. We validate this approach using a custom dataset of long Minecraft videos, where it produces substantially more consistent rollouts compared to existing baselines. --- ### INFUSER: Influence-Guided Self-Evolution Improves Reasoning Authors: Siyu Chen, Miao Lu, Beining Wu, Heejune Sheen, Fengzhuo Zhang, Shuangning Li, Zhiyuan Li, Jose Blanchet, Tianhao Wang, Zhuoran Yang Published: 2026-06-08 Categories: cs.LG, cs.AI, cs.CL, cs.GT, stat.ML arXiv: https://arxiv.org/abs/2606.09052v5 Self-evolution offers a scalable path to stronger reasoning: a pretrained language model improves itself with only minimal external supervision. Yet existing methods either depend on extensively curated or teacher-generated training data, or, when the generator runs unsupervised, reward it by a difficulty heuristic that need not improve the solver. We introduce INFUSER, an iterative co-training framework with two co-evolving roles: a Generator that drafts questions and reference golden answers from a pool of unstructured, automatically collected documents, and a Solver that improves by training on them. The solver is trained with standard correctness rewards against the generator-provided answers, while the generator is rewarded by an optimizer-aware influence score that measures whether each proposed question would actually improve the solver on the target distribution. Because this continuous, noisy influence score is poorly served by standard GRPO, we propose DuGRPO, a dual-normalized variant of GRPO, for generator training. Together, these turn the document pool into an adaptive curriculum that favors questions useful to the current solver, not just hard ones. On Qwen3-8B-Base, INFUSER outperforms strong self-evolution baselines with over 20% relative improvement on Olympiad and SuperGPQA benchmarks, and an 8B INFUSER co-evolving generator outperforms a frozen 32B thinking generator on math and coding. Ablations confirm each design choice is necessary, and two extensions, applying INFUSER to an instruction-finetuned anchor and augmenting it with rule-verifiable RLVR data, further demonstrate the flexibility and generalizability of the framework. Code is available at https://github.com/FFishy-git/INFUSER. --- ### INFUSER: Influence-Guided Self-Evolution Improves Reasoning Authors: Siyu Chen, Miao Lu, Beining Wu, Heejune Sheen, Fengzhuo Zhang, Shuangning Li, Zhiyuan Li, Jose Blanchet, Tianhao Wang, Zhuoran Yang Published: 2026-06-08 Categories: cs.LG, cs.AI, cs.CL, cs.GT, stat.ML arXiv: https://arxiv.org/abs/2606.09052v3 Self-evolution offers a scalable path to stronger reasoning: a pretrained language model improves itself with only minimal external supervision. Yet existing methods either depend on extensively curated or teacher-generated training data, or, when the generator runs unsupervised, reward it by a difficulty heuristic that need not improve the solver. We introduce INFUSER, an iterative co-training framework with two co-evolving roles: a Generator that drafts questions and reference golden answers from a pool of unstructured, automatically collected documents, and a Solver that improves by training on them. The solver is trained with standard correctness rewards against the generator-provided answers, while the generator is rewarded by an optimizer-aware influence score that measures whether each proposed question would actually improve the solver on the target distribution. Because this continuous, noisy influence score is poorly served by standard GRPO, we propose DuGRPO, a dual-normalized variant of GRPO, for generator training. Together, these turn the document pool into an adaptive curriculum that favors questions useful to the current solver, not just hard ones. On Qwen3-8B-Base, INFUSER outperforms strong self-evolution baselines with over 20% relative improvement on Olympiad and SuperGPQA benchmarks, and an 8B INFUSER co-evolving generator outperforms a frozen 32B thinking generator on math and coding. Ablations confirm each design choice is necessary, and two extensions, applying INFUSER to an instruction-finetuned anchor and augmenting it with rule-verifiable RLVR data, further demonstrate the flexibility and generalizability of the framework. Code is available at https://github.com/FFishy-git/INFUSER. --- ### Beyond Convolution: Advancing Hypergraph Neural Networks with Hypergraph U-Nets Authors: Fuli Wang, Wei Qian, Daniel L. Lau, Gonzalo R. Arce Published: 2026-06-08 Categories: cs.LG arXiv: https://arxiv.org/abs/2606.09051v1 Convolutions have successfully transitioned from image processing to the complex realm of non-Euclidean higher-order domains, particularly in hypergraphs. Despite the success in convolution, the exploration of a popular architecture named U-Net remains largely unexplored for hypergraph data due to the lack of well-defined pooling and unpooling operations. This work pioneers the study of U-Net architectures for hypergraph data, addressing the critical challenge of designing effective pooling and unpooling operations that retain maximal structural information from the input hypergraph. Motivated by hierarchical clustering, we propose to construct the pooling and unpooling operators all at once by cutting the clustering dendrogram at different granularities, named the Parallel Hierarchical Pooling (PHPool) and Unpooling (PHUnpool) operators. Unlike existing pooling methods that risk local structural damage through a sequential learning procedure, our PHPool operators are designed in a global and parallel manner to ensure fidelity to the original hypergraph structure with efficient computation while the PHUnpool operators are tailored to perform inverse operations of the PHPools for hypergraph reconstruction. We validate our model through hypergraph reconstruction simulation, hypergraph classification, and node-level anomaly detection, where it demonstrates superior performance over existing state-of-the-art graph and hypergraph deep learning methods. --- ### Data augmented bootstrap: Unifying confidence interval construction by approximate invariance Authors: Kevin Han Huang Published: 2026-06-08 Categories: stat.ME, cs.LG, math.ST, stat.ML arXiv: https://arxiv.org/abs/2606.09049v2 We propose the data augmented bootstrap (DAB), a framework for constructing confidence intervals from approximately invariant transformations of the data. As special cases, DAB recovers popular methods that rely on exact group symmetries, such as conformal prediction, wild bootstrap for Maximum Mean Discrepancy U-statistics and the recently proposed SymmPI. Meanwhile, DAB also recovers the classical bootstrap method, which exploits the dataset's approximate invariance under uniform sampling of data indices as the dataset size grows. For all DAB methods, we establish theoretical coverage results that interpolate between finite-sample and asymptotic guarantees according to the strength of the invariance, and without assuming a group structure. The approximate invariance is measured in the Kolmogorov distance and, for statistics that satisfy Gaussian universality, reduces to conditional mean and variance matching. This allows us to incorporate data augmentation (DA), a widely used machine learning heuristic based on approximate invariances, into known statistical methods. We empirically test the performance of incorporating DA into bootstrap, wild bootstrap and conformal prediction for simulated settings as well as for image, language and scientific data. --- ### BareWave: Waveform-Native Flow-Matching Text-to-Speech Authors: Wei Fan, Chao-Hong Tan, Qian Chen, Wen Wang, Xiangang Li, Kejiang Chen, Weiming Zhang, Nenghai Yu Published: 2026-06-08 Categories: eess.AS, cs.AI, cs.SD arXiv: https://arxiv.org/abs/2606.09048v1 Removing intermediate representations and separately trained decoding stages has become an important direction in generative modeling. In text-to-speech, however, high-quality systems are still commonly built through an intermediate acoustic representation before waveform synthesis. In this work, we present BareWave, a fully waveform-native framework for direct text-to-wave generation in flow-matching TTS. We consider this setting to raise three training challenges: raw-waveform modeling lacks a strong pretrained representational scaffold, different stages of training benefit from different noise schedules, and data-space perceptual objectives do not automatically share the temporal structure of the velocity-space flow objective. As a result, direct waveform training is hard to optimize efficiently, hard to push toward a strong final operating point with a fixed recipe, and hard to integrate effective perceptual refinement. Guided by this view, we develop a direct text-to-wave training framework that combines training-time representation alignment, staged noise scheduling, and velocity-aware perceptual alignment (VAPA), while preserving a single waveform-native inference path without pretrained components at test time. Experiments on zero-shot voice cloning show that strong intelligibility, speaker similarity, and naturalness can be achieved under a fully waveform-native inference path, supporting waveform-native flow-matching TTS as a practical direction. Project page with audio demos is available at https://barewave.github.io/. --- ### Families of Control-Cost-Parametrized Inverse-Optimal Universal Stabilizers Authors: Miroslav Krstic, Luke Bhan Published: 2026-06-08 Categories: eess.SY, cs.LG, math.OC arXiv: https://arxiv.org/abs/2606.09047v1 A classical universal stabilization formula offers the practitioner no design freedom: it is a single, parameter-free object. We introduce a cost-parametrized family of stabilizing feedback laws, where (1) the user chooses a function that serves as the running cost on control in an inverse-optimal cost functional, and (2) obtains, through a formula, a nonlinear "expander" of a pre-existing universal controller, which solves an infinite-horizon optimal control problem with a meaningful cost on the state. The cost-to-expander formula is a three-step construction, involving, inter alia, cost differentiation and function inversion-overall, a nonlinear infinite-dimensional operator. The cost-to-expander operator is proven Lipschitz, which enables uniform neural operator approximation of the entire family and supports both offline performance exploration and online adaptation. Semiglobal practical asymptotic stability and second-order suboptimality bounds are established under the approximation. The operator learning and its use in semiglobal stabilization are illustrated numerically. We call the result 'half-direct-optimal' because the paper's design is less than a general 'direct optimal' (HJB-inducing) control, but more than the fully inverse optimal, since the user performs minimization for an arbitrary given cost on control. The dual to the half-direct problem we solve is the problem in which the cost on the state is arbitrary and given. This dual problem is easier and outside of the scope of the paper. --- ### Measurement Under Selection: Decoy-Calibrated Failure Audits for Language Models Authors: Vyzantinos Repantis, Ameya Gawde, Harshvardhan Singh Published: 2026-06-08 Categories: cs.LG, cs.CL, cs.IR arXiv: https://arxiv.org/abs/2606.09046v2 Knowing how often a language model fails does not explain where its errors concentrate. When auditors examine many explanations, the strongest observed pattern may arise by chance. We introduce Janus, a procedure for checking proposed error patterns before reporting them. Janus starts with a fixed list of yes/no properties of the examples being evaluated, such as whether the input is long. For each property, it compares the model's error rates on examples with that property and those without it. To see how large a difference can arise by chance, it repeats this calculation after shuffling the yes/no labels across examples without changing the group sizes. These shuffled properties are called decoys. A pattern is reported only if the size of its error difference meets a threshold set using decoys. On separate held-out examples, the same group must still have the higher error rate and the difference must meet a minimum, which was chosen in advance. In a controlled experiment, where the model must find a code in documents containing tables of staff, projects, and renewal codes, Janus confirms five related patterns of higher error rates on tasks requiring more lookups across tables. It also confirms a sixth pattern: lower error rates on examples with the needed information at the ends of the tables. In our samples from the MuSiQue and LongBench v2 public benchmarks, SliceLine finds groups with high error rates, while Janus reports no confirmed error patterns for the example properties we chose to test. For comparison, we use standard tests that shuffle errors and account for testing many candidates. With the same holdout check, they confirm two to six controlled patterns, depending on the test and threshold, and none on either benchmark. In simulations with no real error patterns, Janus reports false patterns more often than Benjamini-Hochberg, depending on the decoy count. --- ### Decoy-Calibrated Failure Audits for Language Models Authors: Vyzantinos Repantis, Ameya Gawde, Harshvardhan Singh Published: 2026-06-08 Categories: cs.LG, cs.CL, cs.IR arXiv: https://arxiv.org/abs/2606.09046v1 Useful audits reveal not only how often a model fails, but also where its failures concentrate. An auditor may test many candidate explanations: long inputs, indirect questions, distracting evidence, or combinations of these factors. The risk is selection. The largest observed effect may reflect a real failure mode, or it may simply be the best result among many tried. We introduce Janus, a procedure for deciding when a proposed error explanation is credible enough to report. The goal is not to generate new explanations, but to decide which ones hold up. The auditor starts with a fixed model, a labeled evaluation set, and a frozen list of candidate explanations, which we call descriptors. Janus scores each descriptor by its error-rate lift, then compares real descriptors with fake ones that have the same frequencies but are randomly assigned to examples. A descriptor is confirmed only if it beats this decoy floor on the data used for discovery and then repeats on separate held-out data. In a controlled audit of multi-table lookup tasks, Janus identifies the planted failure, confirming long-chain descriptors and their interactions. The LLM often stops partway through the lookup chain instead of reaching the final answer. On two public benchmarks, MuSiQue and LongBench v2, the SliceLine baseline flags plausible high-error pockets, but Janus confirms none of them. Ablations show why both safeguards matter. On LongBench v2, an uncalibrated fixed threshold reports 20 descriptors, the decoy floor leaves one, and the holdout check rejects the last one after its lift shrinks from 0.36 to 0.05. The resulting principle separates proposing explanations from reporting them. Candidates may come from any source, but only those that beat decoys and replicate on fresh data become audit findings. --- ### DynaCF: Mitigating Shortcut Learning in Reward Models via Dynamic Counterfactual Sensitivity Authors: Fengyuan Liu, Yongliang Miao, Zirui He, Yanguang Liu, Fei Sun, Mengnan Du Published: 2026-06-08 Categories: cs.LG, cs.CL arXiv: https://arxiv.org/abs/2606.09043v1 Reward models trained from pairwise preferences often exploit superficial shortcut cues rather than learning true response quality. We propose DynaCF, a dynamic reweighting framework for mitigating shortcut learning in reward model training. Unlike static shortcut heuristics, DynaCF measures shortcut sensitivity online during optimization by applying semantics-preserving counterfactual perturbations and tracking the resulting margin shifts and preference flips under the current model. Samples with higher shortcut sensitivity are dynamically downweighted in the Bradley-Terry objective, encouraging the model to rely less on superficial patterns and more on task-relevant preference signals. Extensive experiments show that DynaCF consistently improves robustness in preference modeling. --- ### Culturally-Aware AI for Cross-Boundary Community Learning: Undergraduate Innovation at the Intersection of Computation and Design Authors: Jiaojiao Zhao, Weisheng Zhang, Jiawen Cai, Haibin Gao, Luyao Zhang Published: 2026-06-08 Categories: cs.CY, cs.AI, cs.GR, cs.HC, cs.MM arXiv: https://arxiv.org/abs/2606.09041v2 Research on artificial intelligence in education (AIED) is rapidly expanding, yet technical progress often lacks human-centered grounding and adequate attention to cultural context. Community-Based Learning, a pedagogy rooted in social work, remains underrepresented in AIED research, particularly within Asia-Pacific contexts. This paper reports on cross-boundary Community-Based Learning where undergraduate students develop AI-enabled solutions for cultural heritage preservation and sustainable development. We examine how community-engaged computing operationalizes culturally aware, human-centered AIED through participatory elicitation of cultural knowledge, bilingual representation, and stakeholder validation across education, technology, and culture. We contribute a collaborative framework for culturally aware AIED designed to support multi-stakeholder collaboration and widen participation by bridging social work and computational science. --- ### Culturally-Aware AI for Cross-Boundary Community Learning: Undergraduate Innovation at the Intersection of Computation and Design Authors: Jiaojiao Zhao, Weisheng Zhang, Jiawen Cai, Haibin Gao, Luyao Zhang Published: 2026-06-08 Categories: cs.CY, cs.AI, cs.GR, cs.HC, cs.MM arXiv: https://arxiv.org/abs/2606.09041v1 Research on artificial intelligence in education (AIED) is rapidly expanding, yet technical progress often lacks human-centered grounding and adequate attention to cultural context. Community-Based Learning, a pedagogy rooted in social work, remains underrepresented in AIED research, particularly within Asia-Pacific contexts. This paper reports on cross-boundary Community-Based Learning where undergraduate students develop AI-enabled solutions for cultural heritage preservation and sustainable development. We examine how community-engaged computing operationalizes human-centered AIED across three dimensions: education, technology, and culture. We contribute a collaborative framework for culturally-aware AIED that fosters multi-stakeholder collaboration while widening participation by dissolving disciplinary silos between social work and computational science. --- ### Agent Economics: An Entropy-Controlled Pluralistic Alignment Framework for Preventing Artificial Hivemind in Autonomous Agents Authors: Cheonsu Jeong Published: 2026-06-08 Categories: cs.AI arXiv: https://arxiv.org/abs/2606.09039v2 This study proposes the Behavioral Protocol Framework (BPF), an entropy-controlled pluralistic alignment framework designed to address two critical challenges in autonomous agent economies: the hivemind effect arising from excessive strategic convergence among agents and the lack of transparency in autonomous decision-making processes. The proposed BPF consists of three core modules: Mentalizing-based Social Intelligence (MbSI) grounded in Theory of Mind (ToM), Pluralistic Alignment (PA), and a Verifiable Execution Kernel (VEK). These modules are organically integrated within a closed-loop architecture that governs the entire lifecycle of agent behavior, from decision-making and execution to verification and feedback. To evaluate the proposed framework, a simulation environment implemented in Python and a Streamlit-based user interface will be developed. Through empirical experimentation, the study aims to examine whether the entropy-control mechanism of the PA module can effectively preserve strategic diversity among agents and mitigate collective convergence, while the VEK module provides a comprehensive and transparent audit trail of the decision-making process. The anticipated results are expected to demonstrate that the proposed framework can simultaneously enhance the stability, efficiency, and trustworthiness of autonomous agent economies. Consequently, this research offers a practical approach for developing robust, transparent, and accountable agent-native economic systems. --- ### Personalization Meets Safety:Mechanisms,Risks,and Mitigations in Personalized LLMs Authors: Yanyan Luo, Xue Han, Ruiqiao Bai, Xin Huang, Yitong Wang, Qian Hu, Qing Wang, Chunxu Zhao, Jie Liu, Cong Geng, Lehao Xing, Pengwei Hu et al. Published: 2026-06-08 Categories: cs.AI arXiv: https://arxiv.org/abs/2606.09038v1 Large Language Models (LLMs) have enabled increasingly personalized interactions by adapting to users' preferences, contexts, and long-term histories. However, the mechanisms that enable personalization also expand the safety landscape in ways not systematically addressed by existing literature. Existing reviews typically focus either on personalization or safety, leaving their intersection largely unexplored. We present the first comprehensive, safety-aware review of personalized LLMs. We organize personalization along three dimensions-user representation, personalization paradigm, and evaluation-and introduce a unified taxonomy of safety risks. At the representation level, we analyze risks arising from diverse user representations. Across mainstream personalization paradigms, we delineate vulnerabilities inherent to prompting, retrieval augmentation, parameter fine-tuning, reinforcement learning, Mixture-of-Experts (MoE), pruning, agent frameworks, and multimodal personalization, and synthesize mitigation strategies across the model lifecycle. Beyond these fine-grained risks, we characterize paradigm-agnostic safety risks arising from personalized adaptation. We further summarize personalized datasets and evaluation methodologies. Through a case study of OpenClaw, we analyze deployment trends in personalized agent ecosystems. Our analysis reveals three structural inadequacies in existing research: safety is evaluated as user-invariant rather than relational, personalization techniques are analyzed in isolation rather than in composition, and evaluation frameworks cannot capture emergent long-term risks. By jointly examining personalized representations, personalization paradigms, safety risks, defenses, and evaluation methods, we provide a unified framework for developing safe personalized LLMs and highlight key directions for future research. --- ### A Multi-Agent System for Motor Design Optimization via an FEA-AI Hybrid Approach Authors: Jinseong Han, Sunwoong Yang, Namwoo Kang Published: 2026-06-08 Categories: cs.AI, cs.MA arXiv: https://arxiv.org/abs/2606.09037v2 This study presents a large language model (LLM)-based multi-agent framework for interior permanent magnet synchronous motor (IPMSM) design optimization that mitigates limitations of conventional workflows: expertise-dependent problem setup and data preparation, the prohibitive computational cost of finite element analysis (FEA), and the unreliability of AI surrogates in unexplored regions. To this end, we first introduce a Design agent that formulates the optimization problem in natural language, leveraging retrieval-augmented generation to improve answer accuracy on motor design problems from below 50% to 67-80%. Furthermore, a Training agent autonomously repairs improperly defined design spaces by reasoning over solver failure history, raising the success ratio of the geometry sampling from 28% to 84% for AI training. Additionally, to resolve cost and reliability simultaneously, an Optimization agent employs an uncertainty-aware FEA-AI hybrid model: the AI surrogate is the primary evaluator, and FEA is selectively invoked where predictive uncertainty is high. Under the same FEA budget, this hybrid model achieves up to 44% lower iron loss in single-objective and 22.5% higher hypervolume in multi-objective optimization than conventional FEA-only search. Under the same evaluation budget, it reduces computation time by 52-55% while retaining 90-92% of FEA-only hypervolume. Conversely, AI-only search converges to false optima, leaving half its Pareto designs infeasible. Notably, a controller agent adaptively updates the uncertainty threshold that triggers FEA each round, eliminating manual tuning and achieving 5.8% lower single objective iron loss than with a fixed threshold. These results establish domain specialized LLM agents with uncertainty-aware hybrid evaluation as a reliable, scalable paradigm for simulation-driven design automation. --- ### Leveraging NeRF-Rendered Images for 3D Gaussian Splatting Authors: Mizuki Morikawa, Yuta Shimizu, Chunyu Li, Yusuke Monno, Masatoshi Okutomi Published: 2026-06-08 Categories: cs.CV arXiv: https://arxiv.org/abs/2606.09034v1 Neural radiance field (NeRF) and 3D Gaussian splatting (3DGS) are two mainstream approaches for novel view synthesis. They often show complementary performance, i.e., 3DGS demonstrating faster rendering speed and NeRF demonstrating higher rendering quality. Motivated by this, we propose leveraging NeRF-rendered images for 3DGS. Specifically, we target street scenes and utilize a pre-trained street-specific NeRF method to produce training images for a target 3DGS method. In our 3DGS training, NeRF-rendered images are used to remove transient objects in street-level input views and to generate bird's-eye views as additional views, inheriting the higher-quality rendering of NeRF into 3DGS. We further incorporate a diffusion-based image enhancement to improve the image quality of the additional views. Experimental results on one synthetic and two real datasets demonstrate that our proposed method improves street-scene rendering while preserving the speed of 3DGS and the quality of NeRF. --- ### CRANE: Knowledge Editing for Reasoning MLLMs Authors: Han Huang, Hao Wang, Mengqi Zhang, Shu Wu, Qiang Liu, Liang Wang Published: 2026-06-08 Categories: cs.CV, cs.CL arXiv: https://arxiv.org/abs/2606.09033v1 The emergence of reasoning multimodal large language models (MLLMs), which generate explicit chain-of-thought (CoT) reasoning before producing answers, has introduced a new challenge for knowledge editing: methods that appear successful under traditional metrics (teacher-forcing accuracy up to 100%) can fail severely when the model's reasoning process is examined (Grounded Success as low as 0%). We identify three failure modes: (1) Structural Collapse, where weight-modifying methods destroy the CoT format; (2) Cognitive Dissonance, where the model's reasoning chain actively rejects the injected edit fact based on visual evidence; and (3) Shallow Internalization, where methods succeed on exact queries but fail on rephrase or multi-hop variants. On reasoning MLLMs, these modes interact: methods that generalize (FT, LoRA) trigger format collapse, while methods without deep modification cannot generalize. To expose these failures, we propose a CoT-aware evaluation protocol and construct ReasonEdit-Bench, with conflict stratification, multi-level probes, and multi-hop portability tests. We propose CRANE, a retrieval-augmented framework that requires no per-edit parameter modification. CRANE combines a modality-aware dual-library retrieval system with a two-phase training strategy: Supervised Fine-Tuning (SFT) for structural initialization, followed by GRPO with a Cognitive Routing Reward that trains the model to arbitrate between visual priors and injected edit facts. On ReasonEdit-Bench, CRANE achieves 96.9% Grounded Success on conflict scenarios and 96.9% intermediate entity usage in multi-hop chains, with 97.6% text-locality and 68.1% image-locality Edit Independence. On the out-of-distribution MMEVOKE benchmark, CRANE reaches 87.0% under gold retrieval. --- ### Bridging the Agent-World Gap: Text World Models for LLM-based Agents Authors: Yixia Li, Hongru Wang, Peng Lai, Zhiwen Ruan, He Zhu, Youxin Zhu, Ganlong Zhao, Minda Hu, Yun Chen, Sibei Yang, Peng Li, Jeff Z. Pan et al. Published: 2026-06-08 Categories: cs.CL arXiv: https://arxiv.org/abs/2606.09032v1 Large language model (LLM)-based agents are increasingly used in interactive textual environments, from web navigation and code editing to tool use and long-horizon dialogue. Yet many remain largely reactive, mapping observations to actions without an explicit model of how these environments are structured and evolve. This motivates text world models (TWMs): transition models over textual states that, given a state and a candidate action, predict the resulting webpage, terminal output, API response, or user reply, thereby supporting planning, efficient learning, and principled evaluation. We systematically review text world models for LLM-based agents, organized around a formal framework and the agent lifecycle: (1) Foundations, defining text world models and characterizing them by state representation and grounding domain; (2) Construction, taxonomizing LLM-as-WM and code-as-WM paradigms and reviewing methods for building them; (3) Application, examining how world models support agents at training time through experience synthesis and at inference time through planning, verification, and adaptation; and (4) Evaluation, covering both evaluation of the world model itself and its use as an evaluation environment for agents. We aim to consolidate this rapidly developing area, clarify its design space, and highlight open challenges for future research. --- ### TRIAGE: Dialectical Reasoning for Explainable Risk Prediction on Irregularly Sampled Medical Time Series with LLMs Authors: Hyeongwon Jang, Gyouk Chu, Changhun Kim, Joonhyung Park, Hangyul Yoon, Eunho Yang Published: 2026-06-08 Categories: cs.LG, cs.AI, cs.CL arXiv: https://arxiv.org/abs/2606.09030v1 Clinical early warning systems built on electronic health records, in which clinical observations are recorded as irregularly sampled medical time series (ISMTS), must deliver both calibrated risk scores for patient triage and interpretable rationales that clinicians can verify. Large Language Models (LLMs) have been explored for this task, yet they collapse graded clinical risk into overconfident binary predictions. This risk polarization undermines both calibration and cross-patient comparability. To address this, we propose TRIAGE, a framework that trains an LLM to generate dialectical reasoning over competing clinical outcomes by eliciting outcome-specific rationales. This dialectical formulation mitigates risk polarization, enabling a single LLM to yield continuous risk scores grounded in explicit clinical reasoning. Evaluated on three ISMTS benchmarks, TRIAGE achieves an average AUPRC improvement of 3.3% and reduces calibration error by 81% compared to the competitive baselines. An LLM-as-a-judge assessment further shows that our rationales surpass post-hoc explanations from the baseline by 20% in clinical reasoning quality. The source code is available at https://github.com/HyeongWon-Jang/TRIAGE . --- ### Frequency Decoupled Framework for Screen Content Image Super-Resolution Authors: Xufei Wang, Qicheng Zhang, Qi Wu, Ziyang Gu, Shizhuang Weng Published: 2026-06-08 Categories: cs.CV arXiv: https://arxiv.org/abs/2606.09029v1 Methods based on implicit neural representations have demonstrated superior performance in Screen Content Image Super-Resolution (SCISR) . However, they overlooked the inherent frequency characteristics, leading to suboptimal performance. We propose a frequency decoupled framework (FDF) that rethinks SCISR from a phasor perspective by capturing structured energy in amplitude and relational continuity in phase, and jointly exploiting them with bespoke implicit representations to faithfully recover the regular textures and global configuration of Screen Content Image (SCI). Amplitude-Phase Factorization Network (APFN) first separates images into amplitude and phase streams, where Amplitude Clustering Module (ACM) organizes sparse yet high-energy amplitude responses into representative prototypes for periodic pattern extraction, while Phase Consistency Self-Attention (PCSA) progressively reinforces configuration through continuous consistency propagation. And Oscillation-Anharmonic Implicit Fitting Network (OAIF-Net) integrates periodic and coherent implicit representations for efficient exploitation of the periodic patterns and coherent context embedded in SCI. Experimental results show FDF achieves state-of-the-art SCISR performance at multiple scales across four public SCI datasets. Ablation experiments further demonstrate the effectiveness of each component in extracting and exploiting periodic patterns and coherent context. --- ### ATM: Action-Consistency Transfer Matrix for Diagnosing and Improving Latent World Models Authors: Jiaheng Chen Published: 2026-06-08 Categories: cs.CV, cs.AI, cs.RO arXiv: https://arxiv.org/abs/2606.09028v1 Latent world models are increasingly used for control and goal-conditioned planning, yet assessing whether their learned representations are useful for planning usually requires slow, planner-coupled simulator evaluation with CEM or similar planners. Such evaluation is black-box and model-complexity-dependent: under the same protocol, different world models may require minutes to hours per checkpoint. In this work, we propose ATM, an Action-Consistency Transfer Matrix for diagnosing whether latent transitions preserve action semantics relevant to planning. ATM compares action information in real encoded transitions and model-predicted transitions through lightweight post-hoc probes, producing an interpretable matrix that reveals representation quality, transition-domain inconsistency, and failure modes without simulator rollout. It can also be collapsed into a simple screening score for within-task ranking across checkpoints, variants, and world models. When the true success gap is non-trivial, ATM achieves highly reliable pairwise ranking, while reducing minutes-to-hours CEM evaluation to seconds-level transition analysis, yielding more than 100x speedup in our setup. We further introduce AITS, showing that action-identifiability is not only diagnostic but also a useful training signal for improving downstream planning without changing the planner. --- ### SafeRun: Enabling Determinism in LLM Planning for Running Authors: Meilin Chen, Zepeng Zhai, Jiaxuan Zhao, Yuan Lu Published: 2026-06-08 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2606.09027v1 Large Language Models enable flexible natural-language planning but remain unreliable in determinism-critical domains due to their probabilistic nature. This limitation is especially problematic in running planning, where violating safety rules can lead to safety risks. We propose SafeRun, a framework for deterministic LLM-based planning via a decoupled architecture. SafeRun separates soft interpretation by an LLM from hard constraint enforcement by a deterministic solver, ensuring strict safety constraints while preserving natural-language flexibility. To validate SafeRun, we build a comprehensive benchmark for running planning under realistic physiological and safety constraints. Experiments across five LLMs show that SafeRun achieves 100\% safety score (vs.\ 79.1\% PE average and 97.6\% CodeAct average) while maintaining competitive instruction-following scores. The SafeRun benchmark is publicly available at \href{https://huggingface.co/datasets/zzp-seeker/SafeRun-RunPlanning-Benchmark}{huggingface}. --- ### Structural Grid Descriptors Predict Within-Task Solver Success on ARC-AGI Authors: Ayan Pendharkar Published: 2026-06-08 Categories: cs.LG arXiv: https://arxiv.org/abs/2606.09026v2 We ask whether structural properties of intermediate grid states predict whether a symbolic ARC-AGI solver will succeed, framed as a test of conditional mutual information I(X;Y|task) > 0. Across 44,800 runs spanning two architecturally distinct solvers (beam search and Stochastic DFS), 400 ARC tasks, 28 configurations per solver, and both training and evaluation splits, hand-crafted grid descriptors measured at 50% trajectory completion discriminate successful from failed runs within the same task (mean within-task best-feature AUC = 0.885, p < 0.001 under within-task label permutation). Most predictive content lies along a single grid-complexity axis. The result generalizes across solver architectures: a feature selected on one solver predicts success on the other with AUC 0.747-0.762 in all four transfer directions (p < 0.001, leakage controlled). On a pre-registered held-out set of 41 reliable tasks, the frozen feature n_components_final achieves AUC = 0.765 (95% CI (0.717, 0.810), p < 0.001), robust under task-clustered bootstrap resampling and cross-solver task collapsing. The signal is not explained by solver capacity (configuration-residualized AUC = 0.927 and 0.896 for beam search and SDFS, p < 0.001) and is only weakly coupled to score trajectories (R^2 approximately 0). Early stopping at 50% completion reduces beam-search compute by 33.6% while retaining 98.9% of solves; degenerate-trajectory detection reduces SDFS compute by 65.3% with no solve loss. Finally, on 229 of 400 evaluation tasks the DSL primitive library produces no valid transition from the input grid. This 0-step collapse is invariant to search budget and universally failed by beam search, indicating a DSL coverage limitation rather than a search-budget effect. --- ### Personal Salience: Highlighting Is Social, but Individuality Lives in Selection Authors: Kazuki Nakayashiki, Keisuke Watanabe Published: 2026-06-08 Categories: cs.IR, cs.CL, cs.HC, cs.SI arXiv: https://arxiv.org/abs/2606.09024v1 Social highlighters let people mark passages that matter to them. We ask how much of an individual is recoverable from these naturalistic traces, using a co-readership identity control (the same document highlighted by many users) that holds document and topic fixed and asks whether a person's own history predicts their marks better than another reader's does. We separate generic salience (structure), crowd salience (what others marked), and personal salience (the individual residual). First, highlighting is social: which sentences you mark is predicted far better by the crowd than by structure or by a personal model, and even a well-estimated crowd, an information-privileged baseline that sees others' marks on the same document, beats a frontier LLM twin built from your other-document history; the within-document personal signal is at most a whisper (own-vs-other gap +0.017 by an embedding scorer, small but significant). Second, in sharp contrast, individuality lives in selection: asked which of the already-salient passages are yours, your own history is a strong, leakage-free predictor (gap +0.14). A topic decomposition shows this is largely stable thematic preference: it shrinks ~6-8x against a topically-matched peer, and a thin residual cannot be separated from finer topic. The non-obvious part is an asymmetry: under the same scorer the individual signal is ~6-8x weaker in salience than in selection. Methodologically, naive history-conditioning evaluations leak (the target's own marks enter the profile in ~42% of pairs, inflating personal scores by up to +0.15 AP) and small crowds overstate personalization; our results are leakage-free, use a dense crowd, and a model-matched control. Highlights carry a genuine individual signature, but a thin layer over a strong shared one, surfacing far more in which salient things a person selects than in what is salient. --- ### TLDR: Compressing Audio Tokens for Efficient Autoregressive Text-to-Speech Authors: Yejin Lee, Junwon Moon, Hyoeun Kim, Hyunjin Choi, Heeseung Kim, Kyuhong Shim Published: 2026-06-08 Categories: cs.SD, cs.AI arXiv: https://arxiv.org/abs/2606.09019v1 Codec-based autoregressive (AR) speech language models have achieved strong text-to-speech (TTS) quality by modeling speech as sequences of discrete audio tokens with large pretrained backbones. However, this token-level formulation creates a structural efficiency bottleneck: speech-token sequences are much longer than text sequences, requiring the AR backbone to perform causal computation at every token position and maintain a KV cache that grows with the sequence length. We introduce TLDR, a patch-based autoregressive framework that accelerates codec-based AR-TTS by shifting the causal modeling from token-level speech sequences to patch-level sequences. TLDR groups consecutive codec tokens into compact latent patches using a lightweight compressor, models the resulting shorter patch sequence with a frozen pretrained AR-TTS backbone adapted by LoRA, and reconstructs fine-grained speech tokens within each patch using a speaker-conditioned extractor. With a patch size of 4, TLDR achieves a 1.8x inference speedup over the baseline AR-TTS model and reduces global KV-cache memory by up to 75%. Experimental results indicate that patch-level global causal modeling can be a practical way to reduce the inference cost of pretrained codec-based AR-TTS systems without replacing the existing modules. --- ### Beyond Averages: Evaluating LLMs on Human Survey Replication at the Distributional Level Authors: Jeonghyeon Moon, Jiwon Kim, Yeheum Lah, Yoonju Han, Yuncheol Kang Published: 2026-06-08 Categories: cs.CL arXiv: https://arxiv.org/abs/2606.09013v2 LLMs are increasingly used to simulate human survey responses, but prior work has mainly evaluated replication using mean-level or aggregate agreement, offering limited insight into whether LLMs reproduce the variability of human behavior. We evaluate LLM-based survey replication at the distributional level using a non-public 2010 consumer choice experiment on Korean instant noodle purchases, a setting unlikely to overlap with model training data. We evaluate three response variables of differing statistical type: binary purchase incidence, categorical brand choice, and count purchase quantity. For each, we compare human and LLM responses at mean-level, pattern, and distributional alignment, and against reference baselines from the human data alone. LLMs reproduce condition-level patterns reasonably well but fail to capture distributional structure: for purchase quantity, no model beats a condition-insensitive baseline that simply matches the pooled human distribution. Because models that match human means well can still produce distributions further from humans than this baseline, mean-based evaluation alone can be actively misleading. Replication also varies with input configuration, with structured personas and multimodal inputs improving alignment while explicit reasoning prompting degrades it monotonically. --- ### Understanding Quantization-Aware Training: Gradients at Quantized Weights Bias to the Low-Loss Basin Authors: Hanyang Li, Jianhao Ma, Ying Cui Published: 2026-06-08 Categories: cs.LG, cs.AI, math.OC, stat.ML arXiv: https://arxiv.org/abs/2606.09012v1 Post-training quantization (PTQ) converts a trained full-precision model into low-bit weights without task-level retraining, while quantization-aware training (QAT) incorporates quantization into the training loop. Although PTQ is efficient and often accurate at moderate bitwidths, it can fail sharply at aggressive bitwidths; QAT is more expensive but can often recover the lost accuracy. We propose a unified geometric framework that explains both PTQ failure and QAT recovery. We model full-precision training as following a low-loss \emph{river} inside a wider \emph{valley}: a normal neighborhood of the river forms a nearly flat \emph{basin}, while leaving this basin incurs a sharp loss increase. When the quantization grid is comparable to the basin width, local PTQ objectives, including rounding and Hessian-based second-order reconstruction, can select a high-loss deployed quantized point outside the basin even when nearby low-loss quantized points exist. In this regime, straight-through-estimator-based QAT has a useful bias: it evaluates gradients at the deployed quantized weights while updating latent full-precision weights, causing the gradient to sense the valley wall and acquire an inward component that steers subsequent quantized iterates back into the basin. We formalize this mechanism through a local landscape model, construct a geometric PTQ failure mode, and prove finite-time QAT recovery under local quantizer-compatibility assumptions. Experiments across vision and language models under multiple neural-network quantization schemes corroborate the predicted basin-crossing failure of PTQ and the corresponding recovery mechanism of QAT. --- ### Scaling by Diversified Experience for Vision-Language-Action Models Authors: Leiyu Wang, Zhaofengnian Wang, Xueqi Li, Luoyi Fan, Cewu Lu, Nanyang Ye Published: 2026-06-08 Categories: cs.CV arXiv: https://arxiv.org/abs/2606.09009v1 Vision-Language-Action models face significant challenges in real-world deployment due to the entanglement of high-level reasoning with low-level control, and the instability of policy optimization. In this paper, we introduce SyVLA, a robust VLA model trained with diversified experiences. We propose an Intention Decoupling algorithm to isolate control-relevant features from reasoning contexts and a similar-sample guided RL pipeline to stabilize policy updates and mitigate distribution shift. Extensive experiments on real-world robotic tasks and multi-modal benchmarks demonstrate that SyVLA achieves superior task success rates and stronger out-of-distribution generalization compared to existing methods, while effectively preserving core vision-language capabilities. Codes and Datasets is released on \href{https://sy-vla.github.io/}{project page}. --- ### Sustainability and Artificial Intelligence: Necessary, Challenging, and Promising Intersections Authors: Han-Teng Liao, Zijia Wang Published: 2026-06-08 Categories: cs.SI, cs.AI, cs.CY, cs.ET arXiv: https://arxiv.org/abs/2606.09006v1 Both digital economy and digital technology researchers increasingly recognize the need to better address the role that artificial intelligence (AI) plays in shaping the evolution of the environmental, social and governance aspects of development. It appears that sustainability and AI research converge on the features of wicked problems that are complex, interconnected and dynamic. Building off such convergence, this article aims to map out the necessary, challenging, and promising intersections by providing an overview of the state of art research. Based on 541 bibliographic data collected from the Web of Science (WoS) database, the findings reveal the increasingly central body of work on green and sustainable science and technology in bridging various disciplines, main journals and key topics and concepts. The findings reveal how such interactions can be necessary, challenging, and promising. The article concludes with few general arguments regarding how to diversify and expand the community of practice regarding AI for sustainable development, especially in the areas of expected AI application areas and institutions. --- ### Document-Authored Control-Signal Impersonation: A Low-Cost Indirect Prompt Attack on RAG Safety Boundaries Authors: Jianguo Zhu Published: 2026-06-08 Categories: cs.CR, cs.CL arXiv: https://arxiv.org/abs/2606.09005v1 Retrieval-augmented generation (RAG) systems often serialize user queries, retrieved documents, metadata, system labels, and task instructions into one natural-language prompt. We study a source-authority boundary failure in this design: attacker-authored retrieved text can impersonate metadata, provenance, authority, or disclosure-policy signals that appear control-relevant to the model. We call this pattern Document-Authored Control-Signal Impersonation (DACSI). DACSI is a non-imperative, metadata-like payload subclass within indirect prompt injection. Its central lesson is simple: document-authored labels are data, not policy. Command-style injection asks the model to ignore, override, or violate policy; DACSI asks whether untrusted document text can be misattributed as an authorized control signal when RAG prompt rendering collapses trusted and untrusted text into the same natural-language channel. We evaluate DACSI across six model settings, prompt-pressure levels, injection baselines, signal taxonomies, RAG-mediated pipelines, system-control probes, a source-authority attribution probe, and synthetic canary formats. We interpret the evidence by model regime rather than as six equal replications: DeepSeek V4 Pro and Qwen3.5-397B provide the cleanest positive lift, DeepSeek V4 Flash is a high-susceptibility setting, GPT-5.5 and Gemini 3.1 Pro Low are strong-boundary probes with selected residual risks, and GLM-4.7 is a saturated leakage boundary case. Across these regimes, DACSI warrants separate evaluation because it uses a command-free metadata/provenance/policy surface, follows a RAG-specific source-authority path, and responds to source/channel separation. The source-authority probe is behavioral attribution evidence, not proof of an internal mechanism. --- ### LATTEArena: An Evaluation Framework for LLM-powered Tabular Feature Engineering (Extended Version) Authors: Ankai Hao, Ke Chen, Huan Li, Lidan Shou Published: 2026-06-08 Categories: cs.AI arXiv: https://arxiv.org/abs/2606.09004v2 Feature engineering remains a cornerstone of tabular data analysis, and Large Language Models (LLMs) have emerged as a promising paradigm for its automation, giving rise to LLM-powered Automated Tabular Feature Engineering (LATTE). However, the field lacks standardized, cost-aware evaluation platforms, and the combinatorial explosion of design choices obscures true algorithmic progress. To bridge these gaps, we systematically deconstruct 15 representative LATTE methods into a unified 6-dimensional taxonomy. Based on this abstraction, we introduce LATTEArena, a standardized, modular, and extensible benchmarking framework that decouples monolithic pipelines into reusable execution blocks. By distilling the massive combinatorial space, we evaluate 24 core LATTE configurations across 7 research questions. Our head-to-head benchmarking goes beyond predictive accuracy to quantify token efficiency and execution robustness, yielding 17 empirical findings on cost-effectiveness trade-offs. Furthermore, we provide 3 concrete recommendations for optimal real-world deployment. By enabling controlled component-level comparisons, LATTEArena shifts the paradigm from ad-hoc prompt engineering to systematic context management. All code, datasets, and over 4,000 execution logs are publicly available to foster a dynamic, community-driven benchmark. Our framework, leaderboard, and all artifacts are hosted on the LATTEArena project website at https://goodenhak.github.io/LATTEArena. --- ### Multi-Armed Bandits with Arriving Arms: Sequential Screening, Dynamic Regret, and Sublinear Guarantees Authors: Deqi Zheng, Xiaoyang Xu, Yuhong Yang Published: 2026-06-08 Categories: stat.ML, cs.LG, math.ST arXiv: https://arxiv.org/abs/2606.09002v1 We study a stochastic multi-armed bandit problem in which the set of available arms expands over time. This setting arises in sequential experimentation when new actions or treatments become available during an ongoing study, making regret against a single best arm in hindsight inappropriate. We instead evaluate performance relative to the best arm currently available, leading to a dynamic-regret criterion for arriving-arm environments. To address the resulting challenges of arrival information discrepancy (AID) and a drifting benchmark (DB), we propose UCB for Arriving Arms (UCB-AA), an elimination-based procedure with an aiding preliminary screening step for newly arrived arms before full competition with incumbent arms. We show that UCB-AA attains regret bounds that depend explicitly on the arrival process, achieves sublinear dynamic regret under regularity conditions on gap evolution, and admits an online extension for unknown horizons. Simulation results show that UCB-AA reduces wasted pulls and maintains a smaller active arm set while preserving competitive regret performance. --- ### The Token Not Taken: Sampling, State, and the Stochasticity of AI Agents Authors: Muhammad Zia Hydari, Raja Iqbal Published: 2026-06-08 Categories: cs.AI, cs.CY, econ.GN arXiv: https://arxiv.org/abs/2606.08998v3 Agentic AI systems can behave differently across runs: the same request may produce a different plan, a different tool call, a different code edit, or a different final answer. Such variability arises from several layers that are often conflated. At the core of many current agents is a foundation model, a large pretrained model adaptable to many downstream tasks, embedded in an orchestration loop that plans, calls tools, observes results, and updates state. One explicit intrinsic source of variability in such systems is token generation: the model computes scores over possible next tokens, the scores are converted into probabilities, and a decoder may sample tokens using a pseudo-random number generator. A small sampled token difference can then cascade downstream into a different tool call, code path, search query, or agent state. Other sources of variability are extrinsic to token sampling, including changing environments, live data, serving infrastructure, batch effects, and numerical details. By separating these layers, this tutorial clarifies what it means to call agentic AI systems stochastic, when such variability can be reproduced under matched conditions, and why deterministic execution need not imply identical behavior in deployed settings. --- ### Language-Aware Token Boosting: LLM Language Confusion Reduction Without Tuning Authors: Trapoom Ukarapol, Pakhapoom Sarapat, Nut Chukamphaeng Published: 2026-06-08 Categories: cs.CL arXiv: https://arxiv.org/abs/2606.08994v1 Large language models (LLMs) sometimes exhibit language confusion when generating non-English text. Existing approaches typically rely on fine-tuning to mitigate this issue. In contrast, we propose a tuning-free paradigm for reducing language confusion. Within this paradigm, we introduce two methods: Language-Aware Token Boosting (LATB), which applies targeted perturbations to tokens associated with the desired language, and Adaptive Language-Aware Token Boosting (Adaptive-LATB), which dynamically adjusts these perturbations based on the model's confidence in the intended language. Experiments demonstrate that our methods effectively improve multilingual alignment by reducing language confusion, while maintain the summarization quality without requiring any additional fine-tuning. Our code is publicly available. https://github.com/scbdatax/genai-datax-language-aware-token-boosting. --- ### LEAF: A Learning-Enabled ADMM Framework for Accelerated Convex Optimization Authors: Binh Nguyen, Trinh Tran, Truong X. Nghiem Published: 2026-06-08 Categories: cs.LG, eess.SY, math.OC arXiv: https://arxiv.org/abs/2606.08993v1 We propose LEAF, a learning-enabled ADMM framework for accelerated convex optimization. The key idea is to approximate the Moreau envelope of the objective function using an Input Convex Neural Network (ICNN), resulting in a learned model that preserves convexity and smoothness. This leads to the proposed Moreau Envelope Learning ADMM (MEL-ADMM) and its splitting variant sMEL-ADMM. Unlike existing approaches that learn high-dimensional operators directly, LEAF learns a scalar-valued Moreau envelope, significantly reducing model complexity and improving data efficiency. The framework accommodates a broad class of convex problems with smooth and non-smooth objectives. By embedding convexity explicitly through the ICNN architecture, the proposed approach maintains high approximation accuracy while preserving key structural properties of the optimization problem. Both MEL-ADMM and sMEL-ADMM are developed with theoretical guarantees of convergence and feasibility under the learned model. Rigorous analysis shows that the proposed methods achieve convergence rates comparable to classical ADMM while reducing per-iteration computational cost. Numerical experiments demonstrate up to an order-of-magnitude speedup over state-of-the-art solvers while maintaining low optimality gaps --- ### SpaceVLN: A Zero-Shot Vision-and-Language Navigation Agent with Online Spatial Cognitive Memory and Reasoning Authors: Yucheng Deng, Pingrui Lai, Xinhai Li, Chenjia Bai, Xiaoheng Deng, Chengnuo Sun, Xuelong Li, Hua Yang Published: 2026-06-08 Categories: cs.RO, cs.AI, cs.CV arXiv: https://arxiv.org/abs/2606.08992v1 Vision-and-Language Navigation in continuous environments requires agents to understand the spatial structure of previously unseen environments in order to follow language instructions. Although foundation models have opened a promising path toward zero-shot navigation without task-specific policy training, many navigators still rely on local visual cues and linear history-based reasoning, overlooking the spatial nature of navigation across explored regions, traversed paths, landmarks, and their spatial relations. In this paper, we propose SpaceVLN, a navigation agent built around Spatial Cognitive Memory and Task-Guided Spatial Reasoning. Specifically, SpaceVLN introduces an efficient stagewise closed-loop framework where planning and execution are organized around verifiable space--landmark stages. During navigation, the agent progressively abstracts explored regions into Spatial Waypoints and dynamically maintains subtask-grounded landmark evidence, forming a hierarchical Spatial Cognitive Memory for progress localization and spatial-relation understanding. Built on this memory, Spatial-CoT integrates task-progress reasoning with spatial perception, analysis, and prediction, enabling Task-Guided Spatial Reasoning for embodied navigation. The unified stage interface enables SpaceVLN to address both Vision-and-Language Navigation and Object-Goal Navigation under a unified zero-shot setting, without task-specific policy training. Across R2R-CE, RxR-CE, GN-Bench, and HM3D-OVON, SpaceVLN achieves state-of-the-art zero-shot performance, and real-robot deployment further validates its applicability. These results highlight Spatial Cognitive Memory and Task-Guided Spatial Reasoning as a practical foundation for stronger embodied navigation agents. --- ### Structure-Aware Modeling of Multiple-Choice Questions Improves Automatic Difficulty Estimation Authors: Gabriel Ortega, Abelino Jiménez, Séverin Lions, Pablo Dartnell Published: 2026-06-08 Categories: cs.CL, cs.LG arXiv: https://arxiv.org/abs/2606.08988v1 Automatic Question Difficulty Estimation (AQDE) holds growing promise for educational assessment because it has the potential to yield difficulty estimates that are competitive with expert judgment, while helping reduce the time and financial burden associated with pilot administrations and scaling to digital testing contexts. Prior AQDE studies report mixed evidence on whether adding distractors as additional text to the question stem and the correct key consistently improves difficulty prediction. We hypothesize that the effectiveness of distractor information depends on its structural representation, and that explicitly modeling distractors as separate components improves difficulty estimation over baselines that omit this information. To address this, we designed controlled architectures that model MCQ components as distinct inputs to isolate the contribution of distractor content and order. Specifically, we represented distractors by encoding each distractor as its own text input and aggregating their representations either with order-aware concatenation (with positional tags) or with an order-invariant summation. We evaluated these architectures using two Chilean datasets (Natural and Social Sciences, 2016-2020; 4,114 multiple-choice questions). Compared to a simpler model that only used the question stem and the key, our best distractor-aware architecture achieved higher predictive performance, reaching R^2 = 0.83 for Natural Sciences and R^2 = 0.71 for Social Sciences items. An order-invariant variant achieved nearly the same accuracy with approximately half as many parameters, offering a favorable accuracy-efficiency trade-off. These results show that structural information (especially distractor content) drives gains in predictive accuracy, supporting the development of efficient, structure-aware models that are computationally viable for large-scale educational applications. --- ### Beyond Neural Collapse: Task-Intrinsic Geometry Governs Neural Representations in Modular Arithmetic Authors: Hu Tan, Kuo Gai, Shihua Zhang Published: 2026-06-08 Categories: cs.LG arXiv: https://arxiv.org/abs/2606.08985v1 While neural collapse (NC) predicts that a $K$-class-balanced classifier should organize terminal representations as a $(K-1)$-dimensional simplex equiangular tight frame (ETF), modular addition consistently enters a different regime: networks compress to a two-dimensional cyclic geometry in which both classifier weights and token embeddings lie on circles. We refine the explanation of this phenomenon in three directions. First, we formalize a layerwise non-uniform training mechanism: downstream classifier weights are driven by dense cross-entropy gradients into a rank-2 equiangular configuration before upstream embeddings fully reorganize, and once this classifier plane forms, backpropagated feature gradients constrain embedding motion to the same plane while weight decay suppresses orthogonal components. Second, after this subspace locking, the induced in-plane dynamics admit an entropy-regularized transport interpretation on $S^1$; combined with modular-addition labels, this reduces embedding formation to phase alignment, whose minimizers are single-frequency characters of $\mathbb{Z}/P\mathbb{Z}$ and hence equal-angle points on a circle. Third, we quantify why this solution prevails over NC: a simplex ETF gains only an $O(1)$ advantage in cross-entropy, whereas the cyclic rank-2 solution enjoys a $Θ(K)$ advantage under Schatten or weight-decay surrogates, yielding a critical threshold $λ_{\mathrm{crit}} = Θ(1/K)$. Our results explain both why classifier weights move first and why embeddings subsequently align with them, showing that grokking on modular arithmetic is governed not by maximal separation alone but by a task-structured trade-off between separation, symmetry, and complexity. --- ### Baichuan-M4: A Clinical-Grade Medical Agent System for Continuous Care Authors: Aiyuan Yang, Canbin Piao, Chengfeng Dou, Da Pan, Dian Wang, Fan Yang, Fei Deng, Fei Li, Guangwei Ai, Hui Liu, Hongda Zhang, Jinyang Tai et al. Published: 2026-06-08 Categories: cs.AI arXiv: https://arxiv.org/abs/2606.08982v2 Baichuan-M4 is Baichuan Intelligence's clinical-grade medical large model, designed for continuous care rather than single-turn medical question answering. It is built as a coordinated medical agent system around three pillars: Baichuan-Harness, a unified runtime that keeps reinforcement-learning training and real-world deployment consistent while enforcing action constraints, tool use, long-term patient memory, and multi-agent coordination; a core reasoning model trained with a continuous-care reinforcement-learning framework that integrates span-level reward modeling (SPAR++), reasoning-path compression, curriculum learning, and stabilized policy optimization; and a clinical tool layer for patient-memory management, authoritative evidence-based retrieval, and multimodal medical perception across documents, X-rays, and dermatology. On a cross-dimensional medical evaluation suite, Baichuan-M4 attains leading results in static medical knowledge and safety, dynamic OSCE-style consultation, long-context clinical memory, evidence-based retrieval, medical document OCR, and multimodal image understanding, while lowering the hallucination rate to 3.3%. --- ### EPS3D: End-to-End Feed-Forward 3D Panoptic Segmentation Authors: Runsong Zhu, Jiaxin Guo, Xiaoyang Guo, Zhengzhe Liu, Ka-Hei Hui, Wei Yin, Kai Chen, Wei Chen, Weiqiang Ren, Yunhui Liu, Pheng-Ann Heng, Chi-Wing Fu Published: 2026-06-08 Categories: cs.CV arXiv: https://arxiv.org/abs/2606.08980v1 This paper introduces EPS3D, a new end-to-end feed-forward framework for open-vocabulary 3D panoptic segmentation. Unlike existing methods relying on additional preprocessing, we design an end-to-end architecture, with a distillation-based training strategy on diverse 3D scenes to predict 3D-aware semantic and instance features from multi-view images, improving 3D consistency and avoiding error accumulation. We further propose a mutual enhancement module to enforce inherent semantic-instance consistency. By aligning semantics within instances (Ins2Sem) and refining instance features with semantic guidance (Sem2Ins), we achieve more coherent 3D scene understanding. Ultimately, EPS3D outperforms SOTA baselines on two benchmarks (e.g., +13% mIoU for semantics on Replica) with high efficiency (e.g., 1s per scene), supporting tasks like robotic manipulation and 3D scene editing. --- ### Heterophily-Aware Adaptive Knowledge Distillation for Hypergraph Neural Networks Authors: Joohee Cho, David Yoon Suk Kang, Yunyong Ko Published: 2026-06-08 Categories: cs.LG arXiv: https://arxiv.org/abs/2606.08978v1 Hypergraph knowledge distillation aims to retain the predictive performance of a hypergraph neural network (HNN) teacher while reducing inference costs through a lightweight student model. In this work, we observe that HNNs exhibit substantially lower prediction performance on heterophilic nodes connected through semantically diverse hyperedges, indicating that the reliability of teacher knowledge varies across nodes. Motivated by this observation, we propose HADES, a heterophily-aware adaptive distillation method for hypergraph neural networks. HADES quantifies node heterophily and leverages it as an estimate of teacher reliability to modulate the transfer of teacher knowledge during distillation. Experimental results on real-world hypergraphs demonstrate that HADES consistently improves student performance across different HNN teachers and distillation objectives. In many cases, the resulting student models surpass the predictive performance of their teachers while achieving up to 12.3 times faster inference. --- ### Online Learning with Recency: Algorithms for Sliding-window Streaming Multi-armed Bandits Authors: Vladimir Braverman, Chen Wang, Liudeng Wang, Samson Zhou Published: 2026-06-08 Categories: cs.LG, cs.DS arXiv: https://arxiv.org/abs/2606.08977v1 Motivated by the recency effect in online learning, we study algorithms for single-pass *sliding-window streaming multi-armed bandits (MABs)* in this paper. In this setting, we are given $n$ arms with unknown sub-Gaussian reward distributions and a parameter $W$. The arms arrive in a single-pass stream, and only the most recent $W$ arms are considered valid. The algorithm is required to perform pure exploration and regret minimization with limited memory, defined as the number of stored arms. The model is a natural extension of the streaming multi-armed bandits model (without the sliding window) that has been extensively studied in recent years. We provide a comprehensive analysis of both the pure exploration and regret minimization problems with the model. For pure exploration, we prove that finding the best arm is hard with sublinear memory while finding an approximate best arm admits an efficient algorithm. For regret minimization, we explore a new notion of regret and give sharp memory-regret trade-offs for any single-pass algorithm. We complement our theoretical results with experiments, demonstrating the trade-offs between sample, regret, and memory. ---