# Research Papers — 2026-06 · bundle page 259 of 260

> Full-text markdown of papers #12901–12950 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/259
[← Previous bundle](https://aipod.dev/papers/archive/2026-06/page/258.md)
[Next bundle →](https://aipod.dev/papers/archive/2026-06/page/260.md)

### Off-the-Shelf LLMs as Process Scorers: Training-Free Alternative to PRMs for Mathematical Reasoning

Authors: Atoosa Chegini, Soheil Feizi
Published: 2026-06-01
Categories: cs.CL, cs.AI, cs.LG
arXiv: https://arxiv.org/abs/2606.01682v1

Selecting the best response from multiple small-model samples using a stronger scorer is a simple inference-time strategy, but fails when the small model has already committed to incorrect reasoning paths. PRM guided search avoids this by scoring candidate continuations during generation, but requires a reward model trained with step-level labels. We propose Chunk-Level Guided Generation, a training-free alternative that uses an off-the-shelf large language model as a process scorer. At each step, a small model samples k fixed-length candidate chunks, while the larger model scores the candidates using likelihoods without generating any text. The selected chunk is committed before the next step, steering generation before errors can propagate. We instantiate this framework with two selection rules: Likelihood-Guided Selection (LGS), which selects the chunk with the highest length-normalized large-model log-probability, and Contrastive-Guided Selection (CGS), which subtracts the small model's log-probability to favor chunks where the large model's preference diverges from the small model's. We show that scoring variable-length reasoning steps with large-model likelihoods is unreliable due to a systematic length bias that persists even after length normalization, and that fixed-length chunks avoid this confound. On GSM8K, MATH, Minerva Math, AMC23, and AIME24 with Qwen2.5-1.5B guided by Qwen2.5-32B and Llama-3.2-1B guided by Llama-3.1-70B, CGS outperforms majority voting by up to 28 pp and, under matched guidance budgets, matches or outperforms Qwen2.5-Math-PRM-72B guided search on most benchmarks without reward-model training. With Qwen2.5-7B guided by Qwen2.5-72B, CGS reaches 81.8% on MATH and 63.6% on Minerva Math at k=16, surpassing majority voting by 4--6 pp. Finally, Chunk-Level Guided Generation produces substantially shorter reasoning traces than PRM guided search.

---

### Don't Let a Few Network Failures Slow the Entire AllReduce

Authors: Peiqing Chen, Jiedong Jiang, Nengneng Yu, Yuefeng Wang, Sixian Xiong, Wei Wang, Zaoxing Liu
Published: 2026-06-01
Categories: cs.DC, cs.LG, cs.NI
arXiv: https://arxiv.org/abs/2606.01680v1

Network failures are among the most frequent hardware faults in large-scale GPU clusters and a leading cause of training-job interruptions. Modern collective communication libraries such as NCCL mitigate network failures by rerouting traffic through surviving NICs on the same server, trading reduced inter-node bandwidth for uninterrupted training. However, the degraded server remains on the critical path of the standard ring algorithm, slowing the entire collective. We present the first information-theoretic lower bound on AllReduce completion time under asymmetric network bandwidth and show that when the straggler retains at least half of its original bandwidth, the unavoidable overhead relative to the fault-free optimum is only O(1/p) for p GPUs. We then design OptCC, a four-stage pipelined AllReduce algorithm that approaches this lower bound. Experiments on SimAI confirm that OptCC closes the gap left by existing fault-tolerant schemes: under practical network failures with up to 50% bandwidth loss, OptCC completes AllReduce within 2-6% of NCCL's fault-free ring performance, whereas the state-of-the-art incurs up to 57% overhead.

---

### Encoded but Not Routed: Explaining the Table-Chart Gap in Scientific Claim Verification

Authors: Sunisth Kumar, Xanh Ho, Tim Schopf, Andre Greiner-Petter, Florian Boudin, Akiko Aizawa
Published: 2026-06-01
Categories: cs.CL
arXiv: https://arxiv.org/abs/2606.01679v1

Multimodal LLMs are increasingly used to assist scientific peer review, where a core requirement is verifying whether claims in a paper are supported by its evidence. Prior work has shown that models perform substantially better at this task when the evidence is a table than when it is a chart of the same underlying data. This raises the question of whether models fail to extract information from charts, or do they extract it but fail to use it when forming their prediction? We study this question through layer-wise linear probing and attention analysis on three open-weight VLMs over table and chart evidence, representing the same underlying data. We find consistent evidence for the latter. Chart information is encoded in the models' intermediate representations but does not reach the prediction position, a gap that is absent for tables and holds across all conditions tested. Attention analysis further reveals that this disconnect takes two architecturally distinct forms across model families. These findings reframe the table-chart gap as a failure of how encoded visual information is routed at prediction time, rather than a failure of encoding itself.

---

### Why Do Self-Harm Prediction Models Struggle to Generalise? Lexical and Semantic Variations in Emergency Department Triage Notes

Authors: Liuliu Chen, Mike Conway, Jo Robinson, Vlada Rozova
Published: 2026-06-01
Categories: cs.CL
arXiv: https://arxiv.org/abs/2606.01678v1

Self-harm presentations to emergency departments (EDs) are strongly associated with higher suicide risk. NLP models have shown robust performance in detecting self-harm from triage notes within single hospitals, yet performance often declines across institutions. To examine potential causes, we compare ED triage notes from two hospitals by analyzing lexical characteristics, highly associated predictive features, and salient topics. Our results reveal variation in lexical expression and feature importance related to self-harm across hospitals, despite consistent core themes such as self-poisoning and self-injury. These documentation differences are associated with reduced cross-site performance. Our findings provide insight into how institutional variation affects the identification of self-harm in clinical text and highlight potential methods to improve model generalisability.

---

### RDA: Reward Design Agent for Reinforcement Learning

Authors: Hojoon Lee, Ajay Subramanian, Ben Abbatematteo, Vijay Veerabadran, Pedro Matias, Karl Ridgeway, Nitin Kamra
Published: 2026-06-01
Categories: cs.LG
arXiv: https://arxiv.org/abs/2606.01672v1

Reinforcement learning has enabled the acquisition of impressive robotic skills, but typically requires hand-crafted reward functions that are slow to design and difficult to align with human intentions. Recent work, such as Eureka, automates reward design by using an LLM to iteratively generate and refine reward code from task descriptions. However, they rely on coarse feedback signals such as success rate, which provide little semantic insight into the learned behavior. As a result, their trained policies achieve the final goal but are frequently poorly aligned with task instructions. We introduce the Reward Design Agent (RDA), a VLM-based agentic framework that injects semantic understanding into reward design. RDA decomposes tasks, visually evaluates trajectories, summarizes failure modes, and iteratively revises reward code to better align with task instructions. Across 12 tabletop manipulation tasks from ManiSkill and 4 whole-body manipulation tasks from HumanoidBench, RDA produces policies substantially more instruction-aligned than those of other baselines, while achieving comparable task success rates. Videos and the generated reward code are available on https://nitinkamra1992.github.io/reward-design-agent.

---

### When Meaning Travels: A Granular Lens on Hybrid-MoE's Role in Idiomatic Understanding for Language Models

Authors: Sarmistha Das, Vaibhav Vishal, Shreyas Guha, Amaan Ali, Kitsuchart Pasupa, Sriparna Saha
Published: 2026-06-01
Categories: cs.CL
arXiv: https://arxiv.org/abs/2606.01671v1

In the contemporary epoch of multilingual education, learning idioms provides a fascinating gateway towards creativity, cultural values, historical context, and diverse perspectives inherent to various linguistic traditions. This paper showcases the navigation of retaining figurative and cultural semantics in low-resource Southeast Asian languages such as Hindi, Bengali, and Thai, where culturally rich idioms pose significant obstacles for computational modeling and cross-linguistic transfer due to their deep metaphorical complexity. To tackle such complexity, we present Varnika, a reconstructed multimodal idiom corpus comprising 3,533 multilingual idioms, enriched with seven idiomatic tones aligned with both textual and visual representations. Additionally, to infer informative idiomatic understanding, we introduce a Hybrid Mixture-of-Experts (HybridMoE) framework that embeds multiple idiomatic expert opinions while mitigating expert sparsity by integrating outputs from both selected and unselected experts through controlled hybridization, further augmented with Idiomatic Property Signals via masked multimodal embeddings. To analyze the performance across multiple dimensions, we propose the IDIO-TONE and Idiomatic Validation Score, a three-stage evaluation pipeline measuring (i) literal translation fidelity, (ii) visual-semantic alignment, and (iii) idiomatic meaning retention. Empirical evaluations highlight that HybridMoE achieves 5--6\% performance gains across advanced vision language models, demonstrating improved representation of figurative language and culturally embedded meaning in multilingual multimodal settings

---

### Time-Aware Diffusion based on Preference Disentanglement for Generative Recommendation

Authors: Bangguo Zhu, Peng Huo, Yuanbo Zhao, Zhicheng Du, Jun Yin, Senzhang Wang
Published: 2026-06-01
Categories: cs.IR, cs.AI
arXiv: https://arxiv.org/abs/2606.01670v3

Recently, Generative Recommenders (GRs) have emerged as a transformative recommendation paradigm by replacing traditional item IDs with semantic indices (SIDs). Owing to the exceptional generative capabilities of diffusion models, a few pioneering works explore developing GRs with diffusion architectures as the backbone. However, a fatal limitation of existing diffusion-based GRs is that the diffusion process applies uniformly to all items within the historical interactions. In contrast, the user preference is shaped by multifaceted time-evolving factors and thus exhibits a non-stationary distribution in the temporal aspect. To bridge this gap, this study proposes a novel GR framework, named TDPM, by designing the time-aware diffusion on SID tokens. Specifically, TDPM explicitly integrates the impact of time-evolving user preferences into the diffusion process. In detail, the user preference is disentangled into (i) the period preference, which remains consistent over a long time-span, and (ii) the point preference, which is triggered by recent focal events. Extensive experiments on three public real-world datasets demonstrate the significant superiority of TDPM over the state-of-the-art baselines. TDPM achieves average improvements of up to 29.21% and 25.45% in terms of HR@20 and NDCG@20, respectively. The ablation study further underscores the necessity of time-aware token diffusion in diffusion-based GRs.

---

### Time-Aware Diffusion based on Preference Disentanglement for Generative Recommendation

Authors: Bangguo Zhu, Peng Huo, Yuanbo Zhao, Zhicheng Du, Jun Yin, Senzhang Wang
Published: 2026-06-01
Categories: cs.IR, cs.AI
arXiv: https://arxiv.org/abs/2606.01670v1

Recently, Generative Recommenders (GRs) have emerged as a transformative recommendation paradigm by replacing traditional item IDs with semantic indices (SIDs). Owing to the exceptional generative capabilities of diffusion models, a few pioneering works explore developing GRs with diffusion architectures as the backbone. However, a fatal limitation of existing diffusion-based GRs is that the diffusion process applies uniformly to all items within the historical interactions. In contrast, the user preference is shaped by multifaceted time-evolving factors and thus exhibits a non-stationary distribution in the temporal aspect. To bridge this gap, this study proposes a novel GR framework, named TDPM, by designing the time-aware diffusion on SID tokens. Specifically, TDPM explicitly integrates the impact of time-evolving user preferences into the diffusion process. In detail, the user preference is disentangled into (i) the period preference, which remains consistent over a long time-span, and (ii) the point preference, which is triggered by recent focal events. Extensive experiments on three public real-world datasets demonstrate the significant superiority of TDPM over the state-of-the-art baselines. TDPM achieves average improvements of up to 29.21% and 25.45% in terms of HR@20 and NDCG@20, respectively. The ablation study further underscores the necessity of time-aware token diffusion in diffusion-based GRs.

---

### ATLAS: Agentic Test-time Learning-to-Allocate Scaling

Authors: Peijia Qin, Qi Cao, Pengtao Xie
Published: 2026-06-01
Categories: cs.LG
arXiv: https://arxiv.org/abs/2606.01667v1

Test-time scaling has become a major way to improve large language model reasoning, but its orchestration has remained designer-engineered: a fixed sample budget, a fixed refinement loop, a fixed scoring rule, or a fixed search policy decides how compute is spent, leaving the model in charge of solving but not of orchestration. We introduce ATLAS, an agentic test-time scaling framework in which an LLM orchestrator owns the control loop end-to-end. Through a single action, explore, which dispatches a fresh independent solver on the original problem, the orchestrator decides whether to gather more evidence, when to stop, and how to synthesize the final answer; the action space is extensible, with each explore call optionally specifying solver, reasoning effort, or prompting strategy. We evaluate ATLAS on four benchmarks covering scientific question answering, code generation, and multimodal reasoning under a Claude Sonnet 4.6 backbone, where it reaches 56.00% on HLE-Verified, 82.29% on LiveCodeBench, 85.75% on GPQA-Diamond, and 23.71% on BabyVision while using far fewer API calls than fixed-workflow baselines. A multi-model extension, ATLAS-MM, that exposes solver choice as an additional action dimension further improves HLE-Verified to 60.00% and LiveCodeBench to 85.63%, with consistent gains on GPQA-Diamond and BabyVision. Ablations replacing the orchestrator's direct synthesis with a separate integrator degrade or fail to improve accuracy on three of four benchmarks, consistent with the role of stateful evidence management in producing the gains.

---

### DOT-MoE: Differentiable Optimal Transport for MoEfication

Authors: Udbhav Bamba, Arnav Chavan, Aryamaan Thakur, Steve Teig, Deepak Gupta
Published: 2026-06-01
Categories: cs.LG, cs.AI
arXiv: https://arxiv.org/abs/2606.01666v1

The scaling of Large Language Models (LLMs) has driven significant performance gains but created substantial challenges in inference efficiency. While Mixture of Experts (MoEs) architectures address this by decoupling model size from inference cost, training MoEs from scratch is often unstable and compute intensive. Conversion of pre-trained dense models into sparse MoEs has emerged as an alternative solution; however, existing methods typically rely on heuristic neuron clustering or random splitting to partition the Feed-Forward Network (FFN) into experts. In this work, we propose DOT-MoE, a novel framework that formulates the decomposition of dense layers as a Differentiable Optimal Transport (DOT) problem. Instead of static heuristics, we model neuron assignment as a balanced transport problem, utilizing differentiable Sinkhorn-Knopp iterations to enforce strict expert capacity constraints. Furthermore, we utilize Straight-Through Estimators (STE) to jointly learn the discrete neuron-to-expert assignment and the token-to-expert routing policy end-to-end. Extensive experiments across multiple architectures and benchmarks demonstrate that DOT-MoE significantly outperforms structured pruning, heuristic clustering, and random-split baselines, retaining 90% of the original dense model's performance while reducing active parameters by 50%.

---

### Quantifying the Energy Floor: Direct Measurement and Replay Buffer Bias in SAC-Based HVAC Control on sbsim

Authors: Bo Li, Chen Zhang
Published: 2026-06-01
Categories: cs.LG
arXiv: https://arxiv.org/abs/2606.01665v1

We quantify the energy floor -- the minimum achievable cost given action space constraints -- for Soft Actor-Critic (SAC) HVAC control on the sbsim calibrated building simulator. Through minimum-action experiments, we directly measure this floor at USD 35.51/day, dominated by continuous electrical loads (USD 35.44, 99.8%) with negligible gas consumption. The standard SAC baseline, initialized with schedule-policy replay buffer transitions, converges to USD 37.18/day, 4.7% above the floor. We identify buffer initialization as the dominant source of sub-optimality in this scenario: training from an empty buffer reduces cost to USD 35.57/day, eliminating 96% of the gap. Expanding the supply water temperature range by 10 K yields negligible additional savings (USD 0.03/day), and further expansion triggers physical constraint violations. We additionally uncover a discount factor coupling (gamma_eff = 0.891) shrinking the effective planning horizon from 8.3 h to 46 min -- a benchmark-wide issue warranting audit. Systematic ablation across planning horizon, reward weights, and observation enrichment confirms all pre-filled-buffer configurations cluster within 0.7% (USD 37.18--USD 37.42), demonstrating that equipment minimum power -- not algorithmic design -- imposes the binding constraint.

---

### Gate the Filter, Not the Message: Node-Channel Mixtures for Pre-Propagation GNNs

Authors: Zichao Yue, Zhiru Zhang
Published: 2026-06-01
Categories: cs.LG
arXiv: https://arxiv.org/abs/2606.01660v1

Pre-propagation graph neural networks (PPGNNs) push all graph-dependent computation into a preprocessing step and train only on the resulting dense hop features, which makes them highly scalable. A puzzle in this regime is that more complex hop aggregators do not reliably outperform simpler ones: on many benchmarks, a plain MLP-based aggregator matches or beats hop-attention variants. We revisit this behavior from a graph-filter perspective. Over a precomputed diffusion basis, existing PPGNNs differ mainly in how filter coefficients are shared across nodes and feature channels, rather than simply in raw aggregator capacity. MLP-based architectures learn channel-dependent filters that are largely shared across nodes, while hop-attention-based architectures learn node-dependent mixtures that are largely shared across channels. This reveals a missing regime in standard PPGNN designs: joint node- and channel-adaptive filtering under the pre-propagation computational contract. We propose FilterMoE, a mixture-of-experts PPGNN in which a small bank of learnable Chebyshev filter experts is routed jointly over nodes and channels by a 3D gating tensor. Across eleven homophilic and heterophilic benchmarks, FilterMoE outperforms strong PPGNN baselines on nine datasets and ranks first on all three large-scale benchmarks, improving the average test score by 1.53 points. These results establish joint node-channel filter routing as a robust alternative to dataset-specific hop-aggregator selection.

---

### Data-Automated Policy Learning for Nonlinear Welfare

Authors: Chunrong Ai, Zeqi Wu, Zheng Zhang
Published: 2026-06-01
Categories: econ.EM, stat.ME, stat.ML
arXiv: https://arxiv.org/abs/2606.01659v1

This paper explores policy learning from observational data, focusing on a nonlinear welfare criterion in a binary treatment setting. The nonlinear criterion is inspired by scenarios where policymakers prioritize specific population segments. We model this criterion using a utility function that encompasses potential outcomes and intermediate parameters, with the latter capturing higher moments of the outcome distributions. When formulated in the context of observational data, both the intermediate parameters and the welfare criterion depend on the propensity score, which we estimate using machine-learning techniques. To address bias in machine learning estimates, we introduce a novel reweighting-based debiasing approach that offers a promising alternative to traditional orthogonality-based methods. To tackle the complexities of infinite-dimensional policy spaces, we employ sieve approximations and $K$-fold cross-validation for model selection, thereby fully automating the policy-learning process. Despite these complexities, we demonstrate that both the welfare regret and the average welfare regret of our proposed policy learning method satisfy an oracle inequality, thereby providing theoretical guarantees on the performance of the estimated policy relative to the best possible policy. This finding extends the existing results from linear to nonlinear welfare criteria, from finite-dimensional to infinite-dimensional policy spaces, and from a known propensity score to a machine-learned one.

---

### MINTS: Minimalist Thompson Sampling

Authors: Kaizheng Wang
Published: 2026-06-01
Categories: math.OC, cs.AI, cs.LG, stat.ML
arXiv: https://arxiv.org/abs/2606.01655v1

The Bayesian paradigm offers principled tools for sequential decision-making under uncertainty, but its reliance on a probabilistic model for all parameters can hinder the incorporation of complex structural constraints. We introduce a minimalist Bayesian framework that places a prior only on the location of the optimum, while eliminating nuisance parameters through profile likelihood. This yields a generalized posterior that naturally accommodates structural constraints. As a direct instantiation, we develop MINimalist Thompson Sampling (MINTS). For multi-armed bandits with mean constraints, we establish near-optimal non-asymptotic regret guarantees and sharp almost-sure asymptotic regret characterizations. In particular, MINTS attains the classical Lai--Robbins constant in the unstructured setting and automatically adapts to unimodal structure, achieving the sharp constant determined only by the immediate neighbors of the optimal arm.

---

### Physics-Aware Linearized ADMM and Its Unrolling

Authors: Satoshi Takabe, Shunta Arai, Tadashi Wadayama
Published: 2026-06-01
Categories: eess.SP, cs.CV
arXiv: https://arxiv.org/abs/2606.01652v2

Recently, partial differential equations (PDEs) have been used to directly model the measurement process in signal processing, although their evaluation is costly. In this paper, we propose a novel alternating direction method of multipliers (ADMM)-based algorithm called physics-aware linearized ADMM (PA-LADMM) for inverse problems from PDE-based measurement processes. The key idea is the linearization of the subproblem with PDEs, leading to a cost-efficient update rule that calls only a PDE solver and its gradient evaluation per iteration. The algorithm has a theoretical convergence guarantee under certain conditions. In addition, we combine it with deep unfolding to unroll the PA-LADMM and train its internal parameters using supervised data. Two distinct experiments, compressed sensing with optical fiber communication and image restoration from noisy anisotropic diffusion, demonstrated the effectiveness of the proposed algorithms.

---

### Physics-Aware Linearized ADMM and Its Unrolling

Authors: Satoshi Takabe, Shunta Arai, Tadashi Wadayama
Published: 2026-06-01
Categories: eess.SP, cs.CV
arXiv: https://arxiv.org/abs/2606.01652v1

Recently, partial differential equations (PDEs) have been used to directly model the measurement process in signal processing, although their evaluation is costly. In this paper, we propose a novel alternating direction method of multipliers (ADMM)-based algorithm called physics-aware linearized ADMM (PA-LADMM) for inverse problems from PDE-based measurement processes. The key idea is the linearization of the subproblem with PDEs, leading to a cost-efficient update rule that calls only a PDE solver and its gradient evaluation per iteration. The algorithm has a theoretical convergence guarantee under certain conditions. In addition, we combine it with deep unfolding to unroll the PA-LADMM and train its internal parameters using supervised data. Two distinct experiments, compressed sensing with optical fiber communication and image restoration from noisy anisotropic diffusion, demonstrated the effectiveness of the proposed algorithms.

---

### Restoring Initial Noise Sensitivity in Text-to-Image Distillation via Geometric Alignment

Authors: Huayang Huang, Ruoyu Wang, Jinhui Zhao, Wei Deng, Daiguo Zhou, Jian Luan, Yu Wu, Ye Zhu
Published: 2026-06-01
Categories: cs.CV
arXiv: https://arxiv.org/abs/2606.01651v1

Generative distillation significantly accelerates text-to-image (T2I) generation by compressing multi-step trajectories into few-step student models while preserving perceptual quality. However, existing methods primarily optimize efficiency and output fidelity, often neglecting critical properties of the original trajectory. In this work, we identify a key missing property: sensitivity to initial noise, whose degradation impairs downstream control methods relying on noise-based optimization and manipulation. We trace this issue to standard distillation objectives that enforce pointwise output alignment, inadvertently flattening the input-output landscape and suppressing the teacher's local geometric structure. To address this, we propose Geometry-Aware Distillation (GAD), a sensitivity-preserving framework that aligns the local functional behavior of teacher and student models. Specifically, GAD matches Jacobian-vector products with respect to input noise, enabling the student to reproduce the teacher's differential response to perturbations. Extensive experiments across multiple T2I paradigms and noise-driven control tasks demonstrate that GAD significantly restores sensitivity and improves diversity while maintaining high visual fidelity. Code is available at https://github.com/Hannah1102/GAD.

---

### PhyScene3D: Physically Consistent Interactive 3D Tabletop Scene Generation

Authors: Weixing Chen, Zhuoqian Feng, Yang Liu, Yexin Zhang, Yifan Wen, Yinghong Liao, Weichao Qiu, Guanbin Li, Liang Lin
Published: 2026-06-01
Categories: cs.CV
arXiv: https://arxiv.org/abs/2606.01649v2

Generating physically consistent 3D tabletop scenes is a fundamental yet underexplored problem for interactive and generalist robotic learning. The challenge stems from dense object hierarchies and irregular affordances. Here, an interactive scene denotes a physically valid, collision-free environment directly loadable into physics simulators. Existing methods, ranging from decoupled symbolic solvers to end-to-end regression models, often suffer from error propagation or overfitting to noisy supervision containing widespread physical violations. To address these limitations, we introduce PhyScene3D, a framework that reformulates generation as a Human-Mimetic Constructive Process. The proposed Cognitive Topological Reasoning Chain (CTRC) factorizes scene synthesis into a sequential, anchor-conditioned process. It employs a 3D AABB-based placement scheme that imposes a strong structural inductive bias. To address imperfect supervision and physical infeasibility, we introduce Physics-Aware Denoising Alignment (PADA). It integrates a differentiable Signed Distance Field (SDF) with Test-Time Optimization (TTO) to project generated scenes onto a physics-feasible manifold while preserving semantic intent. Experiments demonstrate that PhyScene3D outperforms state-of-the-art approaches in both semantic accuracy and physical validity, achieving a 40% reduction in scene-wise collision rate relative to the human-annotated training data.

---

### Self-Regulating Annealing in Heavy-Tailed Diffusion Models

Authors: Keito Wakatsuki, Hideaki Shimazaki
Published: 2026-06-01
Categories: stat.ML, cs.LG
arXiv: https://arxiv.org/abs/2606.01645v1

Diffusion models have emerged as a leading framework for deep generative modeling. While the standard Gaussian formulation is theoretically convenient, its suitability for heavy-tailed datasets remains unclear. To address this, heavy-tailed diffusion models (HTDMs) extend the standard formulation by replacing the Gaussian distribution with a Student's t-distribution, thereby improving tail fidelity on heavy-tailed datasets. Although stochastic differential equation (SDE)-based sampling is possible in HTDMs, it has not been fully explored. In this paper, we propose an SDE-based sampler for HTDMs that explicitly incorporates a state-dependent diffusion coefficient. This state dependence naturally induces a self-regulating annealing mechanism by adaptively modulating the effective noise scale. We theoretically explore this mechanism and experimentally verify its necessity for reproducing samples from a heavy-tailed distribution.

---

### Conditional Collapse in Sign Language Production: A Diagnostic and a Scaling Argument

Authors: Rui Hong, Jana Košecká
Published: 2026-06-01
Categories: cs.CV
arXiv: https://arxiv.org/abs/2606.01643v1

Sign Language Production (SLP) is the task of generating avatar sign language motion from natural language text. The quality of the generated motion is typically evaluated by a motion-space Fréchet distance (FID) and back-translation (BT) BLEU score on benchmarks such as How2Sign. Both metrics can improve substantially while the underlying generator fails to faithfully represent the sign language gestures. In this work we propose to evaluate the generated motion at three independent levels: ($\tau1$) initial-pose conditioning, ($\tau2$) output diversity, and ($\tau3$) target faithfulness. We compute these as pairwise-distance ratios using latent representations of a frozen motion autoencoder (MoAE). We evaluate 14 SLP model checkpoints on the How2Sign dataset, including a re-implemented Neural Sign Actors (NSA), and show that $\tau3$ faithfulness is never attained, while FID varies by nearly two orders of magnitude and is uncorrelated with faithfulness. We show that on the isolated gloss dataset ASL3DWord favorable $\tau3$ can be attained, hence isolating the size of the sentence-level paired-dataset as the bottleneck.

---

### Edge-directed geometric partitioning for versatile video coding

Authors: Xuewei Meng, Xinfeng Zhang, Chuanmin Jia, Xia Li, Shanshe Wang, Siwei Ma
Published: 2026-06-01
Categories: cs.CV
arXiv: https://arxiv.org/abs/2606.01641v1

To improve the coding performance, geometric partition (GEO) was proposed for the upcoming VVC standard. GEO provides 140 partition candidates. The index of optimal GEO mode needs to be signaled explicitly. Considering different structural characteristics of different CUs and the correlation between spatial adjacent blocks and temporal collocated blocks, we propose a GEO mode prediction strategy by constructing a Most Probable Mode (MPM) list to reduce the overhead of GEO index and improve coding efficiency. Based on the observation of the high correlation between the partition mode and object boundaries, an edge-directed geometric partition scheme is proposed to construct the MPM list according to spatio-temporal edge information. The proposed method provides an objective BD-rate gain of 0.58% and 1.00% on average for RA and LDB configurations compared to VTM-6.0. Besides, it also promotes the visual quality of object boundaries.

---

### MobEvolve: An Agentic Self-Evolving Heuristic System for Interpretable Human Mobility Generation

Authors: Junlin He, Yihong Tang, Tong Nie, Ao Qu, Yuebing Liang, Hamzeh Alizadeh, Bang Liu, Wei Ma, Lijun Sun
Published: 2026-06-01
Categories: cs.AI, cs.CL
arXiv: https://arxiv.org/abs/2606.01640v1

Human mobility generation aims to synthesize realistic trip chains for target populations based on individual features. Existing paradigms, including deep generative models, LLM-based methods, and traditional heuristics, struggle to satisfy the complex demands of this task while simultaneously maintaining interpretability, behavioral plausibility, population-level distributional alignment, and inference efficiency. To bridge this gap, we introduce MobEvolve, the first agentic self-evolving heuristic framework for human mobility generation. MobEvolve initializes a behavior-inspired heuristic system and employs an LLM agent to iteratively evolve its internal logic. By diagnosing empirical misalignments and failure cases on a validation set, the agent proposes targeted updates and accumulates evolution memory for cumulative self-improvement. Extensive evaluations on the Singapore and Montreal benchmarks demonstrate that MobEvolve significantly outperforms state-of-the-art deep generative and LLM-based methods in individual trajectory fidelity, population-level distribution alignment, and behavioral plausibility, while preserving interpretability and high inference efficiency.

---

### CanonCGT: Reference-Based Color Grading via Canonical Pivot Representation

Authors: Jinwon Ko, Keunsoo Ko, Chang-Su Kim
Published: 2026-06-01
Categories: cs.CV
arXiv: https://arxiv.org/abs/2606.01638v1

Reference-based color grading aims to reproduce the tonal mood and lighting of a reference while preserving color harmony and scene structure. Existing photorealistic and filter-based methods often produce unstable tone mappings -- over-shifting or inconsistently retaining colors -- leading to unnatural results. We propose CanonCGT, a two-stage framework built on a canonical pivot -- a style-neutral intermediate representation for stable color mapping. The first stage canonicalizes the input by removing intrinsic tonal bias, and the second color-grades it to match the reference style. A dual-phase training scheme, DP-CGT, combines supervised preset learning with self-supervised refinement on unpaired photographs. CanonCGT delivers photorealistic and tonally consistent results across diverse datasets, surpassing state-of-the-art methods in stability and visual fidelity. Our codes are available at \href{https://github.com/Jinwon-Ko/CanonCGT}{https://github.com/Jinwon-Ko/CanonCGT}

---

### Easier to Mislead Than to Correct: Harmful and Beneficial Revision in LLM Conformity

Authors: Jiaming Qu, Lucheng Fu, Yibo Hu
Published: 2026-06-01
Categories: cs.CL, cs.AI
arXiv: https://arxiv.org/abs/2606.01637v2

Large language models are increasingly used in multi-agent systems, where they see and respond to other agents' answers. A key risk is conformity: a model may abandon its own answer simply because others agree on a different one. Prior studies show that LLMs often revise toward a majority answer, but it remains unclear whether these revisions help correct mistakes as often as they introduce new errors. In this paper, we conduct a controlled study in which an LLM first answers a question, then sees simulated peer responses before making a final decision. We manipulate two social cues: consensus structure and authority labels assigned to peers, and measure how they influence beneficial and harmful revisions. Across four open-weight LLMs and seven QA datasets, we find that peer agreement makes it much easier to mislead initially correct models than to correct initially wrong ones. Authority labels make models more likely to choose the endorsed answer, regardless of whether it is correct. More concerningly, generic reasoning interventions such as chain-of-thought and reflection do not reliably reduce harmful revision while preserving beneficial revision. These findings suggest that multi-agent LLM systems should verify peer answers rather than simply aggregate them.

---

### Pave-GRPO: Beyond Instantaneous Guidance through Principled Average Velocity Decomposition

Authors: Pengyang Ling, Jiazi Bu, Yujie Zhou, Yibin Wang, Zhenyu Hu, Zihan Zhang, Yi Jin, Huaian Chen, Yuhang Zang
Published: 2026-06-01
Categories: cs.CV
arXiv: https://arxiv.org/abs/2606.01636v2

Group Relative Policy Optimization(GRPO) has emerged as an effective paradigm for aligning flow-based generative models with human preferences. However, the high cost of group rollouts forces existing methods to use very few denoising steps, resulting in sparse temporal supervision and leaving most intermediate stages without direct reward guidance. To address this, we propose Pave-GRPO, which reformulates the GRPO objective through principled average velocity decomposition. Rather than generating expensive high-step rollouts, we maintain efficient few-step group sampling but decompose each coarse transition into an equivalent ensemble of finer sub-trajectories spanning multiple intermediate timesteps, propagating reward feedback to a denser set of temporal stages for more comprehensive preference alignment. Crucially, this incurs no additional stochastic rollout generation or reward evaluation: the decomposed sub-trajectories are constructed analytically around the observed transitions, requiring only a small number of extra velocity-network evaluations during the policy update. This design offers two benefits: (i) rollout-free horizon expansion: through the direct reuse of few-step group samples and their associated rewards, Pave-GRPO significantly broadens the effective optimization scope under a fixed sampling and reward budget; and (ii) comprehensive temporal supervision: by equivalently decomposing an instantaneous velocity target into a multi-timestep ensemble, it distributes reward signals across more intermediate stages of the denoising process, enabling finer-grained and more thorough preference optimization.

---

### Pave-GRPO: Beyond Instantaneous Guidance through Principled Average Velocity Decomposition

Authors: Pengyang Ling, Jiazi Bu, Yujie Zhou, Yibin Wang, Zhenyu Hu, Zihan Zhang, Yi Jin, Huaian Chen, Yuhang Zang
Published: 2026-06-01
Categories: cs.CV
arXiv: https://arxiv.org/abs/2606.01636v1

Post-training via Group Relative Policy Optimization (GRPO) has emerged as a powerful paradigm for aligning flow-based generative models with human preferences. However, the iterative denoising nature of flow models incurs substantial costs when generating group rollouts for policy-gradient updates, compelling existing methods to train with extremely few denoising steps. This temporal sparsity severely restricts preference optimization: reward feedback can only reach a handful of stages per trajectory, leaving the vast majority of intermediate denoising steps without direct supervision and thus compromising alignment granularity. To address this, we propose Pave-GRPO, which reformulates the GRPO objective through Principled average velocity decomposition. Rather than generating expensive high-step rollouts, we maintain efficient few-step group sampling but decompose each coarse transition into an equivalent ensemble of finer sub-trajectories spanning multiple intermediate timesteps. This propagates reward feedback to a denser set of temporal stages for more comprehensive preference alignment without additional generation cost. This design offers two benefits: (i) zero-cost horizon expansion: through the direct reuse of piece-wise group samples and their associated rewards, Pave-GRPO significantly broadens the effective optimization scope under fixed sampling budgets; and (ii) comprehensive temporal supervision: by equivalently decomposing an instantaneous velocity target into a multi-timestep ensemble, it distributes reward signals across more intermediate stages of the denoising process, enabling finer-grained and more thorough preference optimization. Extensive experiments validate that Pave-GRPO effectively advances preference alignment across different reward settings, offering comprehensive performance enhancement.

---

### AlphaToken: Decoupling Adaptation and Stability for Path-Aware Response Token Valuation in LLM Post-Training

Authors: Liu Qing, Ou Wu, Yi Du
Published: 2026-06-01
Categories: cs.CL, cs.AI
arXiv: https://arxiv.org/abs/2606.01635v1

Token selection is pivotal for effective LLM post-training. However, existing methods mostly rely on local heuristics and rarely formulate token selection as a principled valuation of individual response tokens. We introduce $\textbf{AlphaToken}$, a response token valuation framework that decouples valuation into $\textbf{adaptation}$ (promoting target-task learning) and $\textbf{stability}$ (preserving pre-trained capabilities), and makes each objective $\textbf{path-aware}$ by combining the direct-path signal from local token gradients with the downstream causal-path signal in autoregressive generation. Since retention data are typically unavailable, AlphaToken approximates stability via a $\textbf{Fisher-drift proxy}$ anchored at the pre-trained reference model. For efficient computation, we extend Ghost Dot-Product to token-level valuation. AlphaToken masks low-value response tokens during fine-tuning and preference optimization, concentrating training signals on more valuable positions. Experiments show that AlphaToken improves post-training performance and mitigates catastrophic forgetting.

---

### E4GEN: Event-level Explainable Extreme-Enhanced Time-series Generation

Authors: Lin Jiang, Dahai Yu, Ximiao Li, Guang Wang
Published: 2026-06-01
Categories: cs.LG, cs.AI
arXiv: https://arxiv.org/abs/2606.01634v2

Generating realistic time series is essential for scientific research and real-world applications. However, existing methods often emphasize overall distributional fidelity while failing to faithfully capture extreme events. To advance existing research, we propose E4GEN, an explainable diffusion framework for extreme event-aware time-series generation. E4GEN provides systematic insights into when, what, and how to control extreme-event generation through three key components. First, E-Activator learns the dataset-adaptive extreme-control signal activation step during the denoising process without interfering with regular temporal components, including trend and seasonality. Second, E-Predictor determines what control signal to enforce through Self-Driven Semantic Prediction, where each sample derives its own control signal by inferring latent extreme-event information during generation. It also includes a novel Data-Conditioned Training, Noise-Initiated Sampling mechanism to address the issue of unavailable training labels. Third, E-Control specifies how to control extreme-event generation through a trainable Extreme Control Network, which transforms the semantic control signal into layer-wise signals and injects it into the denoising process. We evaluate E4GEN on six datasets with 17 metrics, and extensive experiments show that E4GEN outperforms state-of-the-art models across multiple dimensions, including overall fidelity, extreme-event fidelity, and downstream utility.

---

### A Framework for Graph-Conditioned Hierarchical Shapley Attribution in Patent Valuation

Authors: Joy Bose
Published: 2026-06-01
Categories: cs.GT, cs.AI
arXiv: https://arxiv.org/abs/2606.01632v1

Estimating the economic contribution of a single patent inside a product that embodies tens of thousands of patents is a long-standing unsolved problem in intellectual property economics. We propose PatentXAI, a framework that treats patent valuation as a problem of explainable AI: given a characteristic function v(S) encoding the revenue achievable by patent subset S, a patent's Shapley value measures its fair share of product profit in a way that satisfies efficiency, symmetry, dummy, and additivity. To make computation tractable we restrict each patent's coalition to its Markov Blanket inside a knowledge graph, grounded in the C-SVE conditional independence theorem (Li et al., 2020). Scaling experiments from n=12 to n=100 patents using Pareto-distributed coverage graphs report median Markov Blanket size of 32.9 percent of n at n=100, with 90th-percentile blanket size of 55.2 percent of n, and runtime of 10 milliseconds per patent. Difference against exact ground truth at n=12 is 0.088; difference against a high-sample Monte Carlo reference at n=100 is 0.062 plus or minus 0.003. A dense-component experiment shows that when 80 percent of patents share one component, the blanket correctly expands to cover that dense cluster, and the difference versus reference falls to 0.039 because the pooled computation becomes more accurate on homogeneous portfolios. Profit allocation proceeds hierarchically: exact Shapley distributes total profit among macro-components, then centrality-weighted Shapley distributes each component budget among covering patents. Estimating v(S) from real data is the primary open problem; we distinguish this from the computational contribution and outline a concrete roadmap for empirical validation using public ETSI, USPTO, and Lens.org datasets.

---

### Benchmarking LLM-as-a-Judge for Long-Form Output Evaluation

Authors: Junjie Chen, Yuxi Dong, Haitao Li, Weihang Su, Yujia Zhou, Min Zhang, Yiqun Liu, Qingyao Ai
Published: 2026-06-01
Categories: cs.CL
arXiv: https://arxiv.org/abs/2606.01629v4

As large language models (LLMs) are increasingly used for long-form generation, reliably evaluating long-form outputs has become a critical challenge. LLM-as-a-judge offers a scalable alternative to human evaluation, yet its reliability in long-form output evaluation remains underexamined: existing meta-evaluation benchmarks focus mainly on short-form outputs. Compared with short-form evaluation, long-form evaluation is not merely a matter of output length; it often requires judges to make more complex document-level assessments of overall organization, task-relevant coverage and depth, cross-section consistency, and scenario-specific quality criteria. In this work, we introduce LongJudgeBench, a comprehensive benchmark for evaluating LLM judges on long-form outputs across diverse real-world scenarios and judging protocols. We systematically evaluate a broad range of LLM judges, covering multiple base models and judging settings. Our results reveal a substantial reliability gap: current LLM judges remain unstable across scenarios, and rubrics or references are helpful but not always sufficient. We hope LongJudgeBench will support future research on more robust, context-aware, and human-aligned LLM-as-a-judge methods. Our code is available at https://github.com/cjj826/LongJudgeBench.

---

### Benchmarking LLM-as-a-Judge for Long-Form Output Evaluation

Authors: Junjie Chen, Yuxi Dong, Haitao Li, Weihang Su, Yujia Zhou, Min Zhang, Yiqun Liu, Qinyao Ai
Published: 2026-06-01
Categories: cs.CL
arXiv: https://arxiv.org/abs/2606.01629v2

As large language models (LLMs) are increasingly used for long-form generation, reliably evaluating long-form outputs has become a critical challenge. LLM-as-a-judge offers a scalable alternative to human evaluation, yet its reliability in long-form output evaluation remains underexamined: existing meta-evaluation benchmarks focus mainly on short-form outputs. Compared with short-form evaluation, long-form evaluation is not merely a matter of output length; it often requires judges to make more complex document-level assessments of overall organization, task-relevant coverage and depth, cross-section consistency, and scenario-specific quality criteria. In this work, we introduce LongJudgeBench, a comprehensive benchmark for evaluating LLM judges on long-form outputs across diverse real-world scenarios and judging protocols. We systematically evaluate a broad range of LLM judges, covering multiple base models and judging settings. Our results reveal a substantial reliability gap: current LLM judges remain unstable across scenarios, and rubrics or references are helpful but not always sufficient. We hope LongJudgeBench will support future research on more robust, context-aware, and human-aligned LLM-as-a-judge methods. Our code is available at https://github.com/cjj826/LongJudgeBench.

---

### Demystifying Multimodal Biomolecular Co-design With Intrinsic Geodesic Coupling

Authors: Keyue Qiu, Xintong Wang, Zhilong Zhang, Hao Zhou, Wei-Ying Ma
Published: 2026-06-01
Categories: q-bio.BM, cs.AI
arXiv: https://arxiv.org/abs/2606.01628v1

Biomolecules such as proteins and small-molecule ligands play a central role in biological systems, arising from the tight interplay between sequence and three-dimensional structure. Recent generative models for biomolecular co-design aim to capture this interplay by jointly modeling coupled modalities. However, existing approaches largely adopt a parallel execution of marginal generative processes, implicitly enforcing fixed synchronous coupling. We argue that a critical but overlooked degree of freedom lies in how these marginal processes are temporally coupled during training and generation, where inappropriate coupling can introduce high-variance supervision and inconsistent intermediate states, affecting modality consistency. To address this, we introduce GeoCoupling, a systematic framework that optimizes for temporal couplings between heterogeneous modalities. Empirical results across structure-based drug design and unconditional protein design demonstrate the learned couplings consistently outperform synchronous and randomly coupled baselines, yielding biomolecules with improved physical validity and diversity.

---

### IMWM: Intuition Models Complement World Models for Latent Planning

Authors: Baoqi Gao, Ruize Han, Miao Wang, Song Wang
Published: 2026-06-01
Categories: cs.LG
arXiv: https://arxiv.org/abs/2606.01626v1

Planning with a learned latent world model is a promising route to control from raw pixels, but a strong world model alone is not enough. We show this experimentally: even with a perfect world model (operationalized by replacing the learned forward predictor with an idealized rollout of the true environment dynamics), a finite-budget sample-based planner still fails on some tasks, indicating that the bottleneck can lie in search rather than in world-model accuracy. Motivated by this gap, we propose IMWM (Intuition Model + World Model), which pairs the world model with an intuition model trained from demonstrations to recognize promising actions. The two models collaborate through three lightweight components: (i) Retrieval Initialization, which initializes the planner's action proposal from a retrieved demonstration; (ii) Hybrid Cost, which combines the intuition score with the world-model rollout cost; and (iii) a Reliability Gate, which adjusts how much the planner trusts intuition in each setting. Across four pixel-based goal-reaching tasks (Two-Room, Reacher, Push-T, and OGBench-Cube), IMWM has higher mean success than the world-model-only planner on all four, with the largest gains on Two-Room (99.2%, +11.5 percentage points) and OGBench-Cube (94.7%, +28.5 percentage points).

---

### What to Test Next: Interpretable Coverage Gap Discovery in Driving VLMs

Authors: Abhishek Aich, Sparsh Garg, Vijay Kumar BG, Turgun Yusuf Kashgari, Manmohan Chandraker
Published: 2026-06-01
Categories: cs.CV, cs.SE
arXiv: https://arxiv.org/abs/2606.01624v2

Driving vision-language models (VLMs) must accurately understand scenes across diverse conditions defined by Operational Design Domains (ODDs), yet verification remains sparse: many slices are missing, making empirical failure rates unreliable. We propose SliceScorer, a deterministic scoring rule for missing-slice recommendation that combines (i) an exposure-based coverage prior to prioritize rare, under-tested regions, and (ii) a neighbor-failure prior that propagates risk from similar tested conditions. SliceScorer is deliberately simple - interpretable, auditable, and conservative - properties essential for safety-critical validation. For stress testing beyond the declared ODD, we embed SliceScorer within SliceNav, an LLM-orchestrated verification pipeline where the model interprets developer queries to select relevant operators (triage, scoring, acquisition, evaluation) and vocabulary extensions, composing verification workflows while keeping all scoring deterministic and auditable. Experiments on three driving VLMs (WiseAD, DriveMM, Cosmos-Reason2-2B) show that SliceNav surfaces high-risk coverage gaps more effectively than prior slice-discovery methods while maintaining diverse recommendations across the condition space. Ablations confirm both scoring components contribute, and qualitative analysis demonstrates end-to-end workflows from developer query to targeted evaluation.

---

### Goal2Pixel: Grounding Goals to Pixels for Vision-Language Navigation

Authors: Muyi Bao, Yuxin Cai, Hang Xu, Zongtai Li, Jinxi He, Jingfan Tang, Chen Lv, Ji Zhang, Yaqi Xie, Wenshan Wang
Published: 2026-06-01
Categories: cs.CV, cs.RO
arXiv: https://arxiv.org/abs/2606.01621v2

Vision-language models (VLMs) have become a common foundation for vision-and-language navigation in continuous environments (VLN-CE). Yet most VLM-based methods cast navigation as low-level action prediction, an interface that is ambiguous, tied to short-horizon motion primitives, and inefficient due to repeated VLM querying. We propose Goal2Pixel, a pure pixel-based paradigm that reformulates VLN-CE as navigable pixel grounding. Rather than predicting actions, Goal2Pixel uses the image plane as a unified spatial interface between VLM reasoning and robot motion: the model predicts a visible navigable pixel to the agent, which is back-projected into a 3D waypoint for forward navigation. For non-forward actions, we append auxiliary directive regions to the image plane, where the left/right/bottom regions are interpreted as turning left, turning right, and stopping, respectively. To enable long-horizon navigation, we propose a visibility-aware keyframe memory for compact and informative history representation. To adapt pretrained VLMs to navigable pixel grounding, we introduce semantic embeddings and coordinate-aware auxiliary losses. Goal2Pixel achieves competitive state-of-the-art performance while requiring fewer VLM inference calls than prior methods. On R2R-CE Val-Unseen it achieves 54.1% SR and 52.5% SPL with just 7.75 VLM calls per episode, 6x fewer than the 46.62 required by direct action prediction at 32.9% SR. The same trend holds on RxR-CE.Project Page: https://baobao0926.github.io/Goal2Pixel/.

---

### Real-Time Generation of Streamable Talking Portrait Video with Reference-Guided Deep Compression VAEs

Authors: Sicheng Xu, Yu Deng, Shoukang Hu, Yichuan Wang, Yizhong Zhang, Zhan Chen, Jiaolong Yang, Baining Guo
Published: 2026-06-01
Categories: cs.CV
arXiv: https://arxiv.org/abs/2606.01620v1

Video diffusion models have significantly advanced portrait video generation, yet their high computational demands limit their use in interactive applications. This work presents a framework for streamable talking portrait video generation conditioned on speech audio and reference images. Designed meticulously for streaming scenarios, it features a causal video VAE for deep latent compression and an autoregressive latent denoising model. Our causal VAE integrates a variable number of reference images as guidance, allowing the network to focus on dynamic information rather than static appearance, thereby enhancing compression efficacy and reconstruction quality. Additionally, we extend the residual auto-encoding paradigm to improve spatial-temporal causality handling in our VAE. The generator is based on a Rectified Flow Transformer architecture and produces video latents in a blockwise auto-regressive manner. Our method enables the real-time generation of high-quality talking portrait videos, achieving speeds significantly faster than baseline models. Furthermore, comprehensive experiments demonstrate that it is on par with or even outperforms these large models in realism, vividness, and video quality.

---

### ReSkill: Reconciling Skill Creation with Policy Optimization in Agentic RL

Authors: Zelin He, Haotian Lin, Boran Han, Wei Zhu, Haoyang Fang, Bernie Wang, Xuan Zhu, Runze Li, Matthew Reimherr
Published: 2026-06-01
Categories: cs.AI, cs.LG, stat.ML
arXiv: https://arxiv.org/abs/2606.01619v2

Agentic reinforcement learning (RL) enables LLM agents to improve continuously from environment rewards, yet the resulting policies do not systematically accumulate reusable strategies that generalize across tasks. Modular skills can provide such reusable strategies, yet existing skill-augmented RL methods decouple skill creation from policy optimization, risking adopting skills that conflict with the evolving policy. Inspired by Anthropic's Skill Creator, we introduce ReSkill, an RL-in-the-loop skill creation framework that reconciles skill evolution with policy learning. ReSkill exploits the group-wise structure of GRPO to naturally embed three mechanisms with only marginal additional overhead: (1) an assertion-driven skill creator that diagnoses failures from past experience and proposes conditional, trigger-based skill revisions; (2) within-group rollout sampling that enables controlled comparison of skill versions, capturing which version best supports the policy's ongoing learning; and (3) Thompson Sampling with adaptive discounting to balance exploration and exploitation in skill version selection as the policy evolves. Across several domains, ReSkill consistently outperforms existing memory and skill-based RL methods, with the largest gains on unseen tasks. Analysis of the skill lifecycle shows skills being automatically created, tested, refined, and pruned as the policy improves, demonstrating reconciled skill-policy co-evolution.

---

### EvoPool: Evolutionary Programmatic Annotation for Label-Efficient Specialized Supervision

Authors: Tianyi Xu, Yaolun Zhang, Xuan Ouyang, Huazheng Wang
Published: 2026-06-01
Categories: cs.CL, cs.AI
arXiv: https://arxiv.org/abs/2606.01617v1

Large language models excel at general tasks but underperform smaller supervised models in specialized, high-stakes domains where training labels are costly. We address this regime with EvoPool, an evolutionary multi-agent framework inspired by Darwinian evolution. Three specialized agents iteratively propose executable annotator code, a small validation set provides a fitness signal, and a deterministic gate keeps only annotators that pass viability, diversity, and marginal-contribution checks across generations. Pool votes are mapped to soft training labels by EvoAgg, a text-aware aggregator combining semantic features with annotator-vote features. The authored pool runs at near-zero per-example cost and is 4500 to 31000x faster than LLM annotation on 100K examples. Across 7 of 8 LLM-weak specialized and complex tasks spanning biomedical relation extraction, legal-clause classification, complex reasoning, and dense multi-label biomedical classification, EvoPool beats the strongest LLM annotation baseline by an average +0.141 macro-F1, peaking at +0.301 on ChemProt and +0.265 on PubMed. Code is available at: https://github.com/tianyi0216/EvoPool

---

### Turing Patterns for Multimedia: Reaction-Diffusion Multi-Modal Fusion for Language-Guided Video Moment Retrieval

Authors: Xiang Fang, Wanlong Fang, Wei Ji, Tat-Seng Chua
Published: 2026-06-01
Categories: cs.CV, cs.MM
arXiv: https://arxiv.org/abs/2606.01615v1

Video-language models are pivotal for tasks such as moment retrieval and highlight detection, yet they often struggle to capture the dynamic, non-linear interactions between temporal video sequences and textual semantics. Existing approaches, relying on static cross-attention or prompt-tuning mechanisms, fail to adaptively model the evolving relationships between modalities, leading to suboptimal alignment and limited generalization. Inspired by systems biology, we propose \textbf{Reaction-Diffusion Multimodal Fusion (RDMF)}, a novel framework that reimagines video-language alignment as a reaction-diffusion (RD) process, drawing on the principles of pattern formation introduced by Alan Turing. In RDMF, video features diffuse across time to capture temporal context, while text-video interactions are modeled as non-linear reactions that amplify relevant features and suppress noise, forming emergent patterns akin to biological systems. Leveraging the Gray-Scott RD model, we design a computationally efficient fusion module that integrates video and text representations, supported by rigorous mathematical analysis of stability and convergence using Turing instability criteria. Our framework is theoretically grounded, employing advanced mathematical tools to ensure stable pattern formation, and is practically viable, incorporating standard components like pretrained encoders and DETR-style heads for moment retrieval and saliency prediction. RDMF represents a pioneering interdisciplinary approach, bridging systems biology and multimedia research to address the limitations of conventional multimodal fusion. Preliminary experiments demonstrate its potential to outperform existing methods in identifying salient video moments, offering a new paradigm for video-language tasks.

---

### TechRAG: Evidence-Gated Multimodal Agentic RAG for Technical Literature Reasoning

Authors: Kanwar Bharat Singh
Published: 2026-06-01
Categories: cs.IR, cs.AI, cs.MA
arXiv: https://arxiv.org/abs/2606.01613v2

This paper presents an agentic multimodal retrieval-augmented generation (RAG) framework for domain-specific literature reasoning, instantiated on a curated corpus of several thousand papers in intelligent tires, vehicle dynamics, vehicle control, sensing, estimation, and machine learning. Unlike conventional single-pass RAG systems, the proposed architecture uses an autonomous, evidence-gated pipeline that classifies query intent, generates separate text and visual query rewrites, performs hybrid text retrieval with FAISS and BM25 followed by cross-encoder reranking, expands evidence through graph-guided chunk traversal over a Neo4j knowledge graph, and retrieves visual document evidence using ColSmol late-interaction embeddings with MUVERA fixed-dimensional encoding, approximate nearest-neighbor search, and MaxSim reranking. The framework scores evidence sufficiency using a 100-point rubric with hybrid rule-based/LLM review, retries retrieval through drift-guarded reformulation, searches external academic databases through optimize--search--vet loops, merges and deduplicates multimodal evidence, verifies citation integrity, and generates cited answers through Planner, Researcher, Writer, and Critic agents with self-correcting revision. Key contributions include: (i) a scalable multimodal retrieval architecture combining text, graph, and visual evidence over 40,000 document pages; (ii) an interpretable evidence sufficiency and retry mechanism; (iii) a multi-agent generation pipeline with evidence mapping and critic-driven revision; (iv) a domain knowledge graph with LLM-based entity extraction, OpenAlex author validation, and intra-corpus citation resolution; and (v) a route-dependent external search architecture for targeted literature expansion. The result is a practical, evidence-gated, multimodal agentic RAG architecture for technical reasoning over specialized research corpora.

---

### Self-Improving Small Object Grounding in LVLMs

Authors: Tianze Yang, Yucheng Shi, Ruitong Sun, Ninghao Liu, Jin Sun
Published: 2026-06-01
Categories: cs.CV, cs.LG
arXiv: https://arxiv.org/abs/2606.01612v1

Can internal attention patterns in Large Vision Language Models (LVLMs) identify reliable small-object boxes without fine-tuning? In this work, we provide an affirmative answer. Attention structure in LVLMs encodes grounding quality-a lightweight IoU regressor trained solely on attention maps achieves strong IoU prediction (Pearson r > 0.67). This regressor powers the regressor-based variant of our Attention-based Candidate Selection (ACS) framework, called ACS-Learned, which selects the best box from multiple sampled candidates to improve object grounding. By analyzing what the regressor learns, we reveal which transformer layers and heads are most critical and derive ACS-Free: a training-free selector that ranks candidates by attention entropy on these discriminative heads, with no learned component at inference. Experiments on COCO and Objects365 demonstrate up to 19% self-improvement on small object localization, with ACS-Free ranking best among all training-free methods, demonstrating that useful attention structure improves both localization reliability and interpretability in LVLMs.

---

### Revisiting Ripple Effects in Knowledge Editing through Pressure-Aware Joint Neighborhood Optimization

Authors: Haoben Huang, Shuxin Liu, Ou Wu, Di Gao
Published: 2026-06-01
Categories: cs.AI
arXiv: https://arxiv.org/abs/2606.01610v1

Single-edit updates in large language models can trigger ripple effects across local knowledge neighborhoods: desirable propagation to related facts and unintended perturbation of preserved ones. Existing methods address these two effects separately, without explicitly modeling their coupling. We challenge this separation through an analysis of ripple responses across typical baselines, identifying two coupled design pressures: editable-side coordination and preserved-side leakage. We propose Joint Neighborhood Optimization (JNO), a new knowledge-editing framework to formalize and jointly address both pressures at the target-planning stage. JNO instantiates this principle through Pressure-Aware Coordination (PAC), which jointly optimizes neighborhood target representations under coupled constraints, and a semantic pre-execution gate that rejects high-risk target plans before parameter execution. Experiments on RippleEdits show JNO improves propagation and preservation metrics by at least 7.0% while preserving cross-backbone editing stability.

---

### Exploiting Semantic and Pixel Representations for Ultra-Low Bitrate Image Compression

Authors: Hao Wei, Yanhui Zhou, Chenyang Ge, Saeed Anwar, Ajmal Mian
Published: 2026-06-01
Categories: cs.CV
arXiv: https://arxiv.org/abs/2606.01608v1

Most existing extreme compression methods fail to achieve an optimal rate-distortion-perception trade-off, as they typically prioritize perceptual fidelity and visual realism over pixel-level accuracy. Consequently, the resulting reconstructions often deviate noticeably from the originals. Ultra-low bitrate image compression is therefore crucial-not only for producing extremely compact representations but also for ensuring that reconstructed images remain semantically coherent and faithful to the source at the pixel level. To this end, we propose SPRDiff, a diffusion-based compression method that fully leverages both semantic and pixel representations, thereby enhancing reconstruction fidelity under ultra-low bitrate constraints. Specifically, we develop a triple-encoder architecture that utilizes high-fidelity features from the pretrained distortion-oriented and semantic-oriented encoders to compensate for the limited representations extracted by the frozen VAE encoder, thereby improving latent compression and entropy modeling. To further enhance the reconstruction fidelity of diffusion models, we introduce a distortion-aware reconstruction module with dual feature extraction. This module not only generates a coarse reconstruction that preserves the main structures, but also provides practical and accurate semantic- and pixel-level conditional signals to guide the diffusion model. Extensive experiments on benchmark datasets demonstrate that our method outperforms state-of-the-art approaches in the rate-distortion-perception tradeoff at extremely low bitrates (below 0.03 bpp), effectively preserving both perceptual quality and pixel-wise fidelity in the reconstructed images. We will release the source code and trained models at https://github.com/cshw2021/SPRDiff.

---

### FedMTFI: Feature Importance Based Optimized Multi Teacher Knowledge Distillation in Heterogeneous Federated Learning Environment

Authors: Nazmus Shakib Shadin, Aaron Cummings, Xinyue Zhang, Bobin Deng
Published: 2026-06-01
Categories: cs.LG, cs.AI
arXiv: https://arxiv.org/abs/2606.01607v1

Federated learning (FL) is a decentralized approach that enables collaborative model training without exposing raw data. Instead of transferring sensitive data, it allows devices to share only model weights, keeping personal data locally and secure. However, in real world settings, the data held by devices is often not evenly distributed and devices mostly differ in computing power and memory capacity. These differences make FL harder to maintain consistent performance across the system. To address these issues, we propose FedMTFI, a novel architecture that combines multi-teacher knowledge distillation (MTKD) with feature importance to improve the FL process in heterogeneous environments. In FedMTFI, clients are clustered based on similar hardware and model types. Each cluster trains a specific model on not independently and identically distributed (non-IID) data. Within a cluster, every client updates that model using only its own local private data. The server then aggregates the locally trained models in each cluster using FedAvg to form multiple prototype models. Then these prototypes serve as teacher models to train a global generalized student model using MTKD. What makes FedMTFI more unique is the integration of Shapley values (SHAP) to emphasize important features during distillation, which enhances both accuracy and interpretability. Experimental results show that FedMTFI achieves higher accuracy than traditional FL algorithms and performs more effectively under non-IID data conditions.

---

### Paving the Way for Point Cloud Video Representation Learning Using A PDE Model

Authors: Zhuoxu Huang, Zhenkun Fan, Jungong Han, Josef Kittler
Published: 2026-06-01
Categories: cs.CV
arXiv: https://arxiv.org/abs/2606.01604v1

Investigating spatial-temporal correlations, specifically how spatial points vary over time, is crucial for understanding point cloud videos. Traditional methods, particularly flow-based techniques, struggle with these correlations due to the unordered spatial arrangement of sequential point cloud data. To address this challenge, we propose a novel approach that regularizes spatial-temporal correlation learning by formulating the problem as a solvable Partial Differential Equation (PDE). While PDEs have long been effective in the physical domain, their application to novel sequential data like point cloud video remains underexplored. Inspired by fluid analysis, we construct a simplified PDE, and the process of solving PDE is guided and refined by a contrastive learning structure between the temporal embeddings and the spatial embeddings. With this extra supervision, our method, named MotionPDE, serves as an effective, plug-and-play enhancement module for existing backbone models, adding minimal computational overhead and parameters. Capitalizing on the contrastive learning process, we delve deeper into the self-supervised capabilities of MotionPDE, yielding promising results that underscore its utility and adaptability in point cloud video data interpretation. The code repo with trained checkpoints will be available at https://github.com/zhh6425/motionpde.git for facilitating future research.

---

### Estimating Mutual Information between Time Series and Temporal Event Sequences Across Diverse Analysis Tasks

Authors: Haoji Hu, Huaqing Mao, Yijun Lin, Xiaowei Jia, Jinwei Zhou, Minoh Jeong, Yao-Yi Chiang
Published: 2026-06-01
Categories: cs.LG, cs.AI, cs.IT
arXiv: https://arxiv.org/abs/2606.01602v3

Pairwise dependence measures such as correlation and causality are fundamental to temporal data mining, yet there is still no principled and robust way to quantify dependence between heterogeneous data types, especially between continuous time series and discrete temporal event sequences. Existing approaches rely on ad hoc transformations or mutual-information estimators that are highly sensitive to quantization, repeated values, and event redundancy, leading to biased or unstable results in practice. We propose a nonparametric mutual information estimator that directly measures the dependence between time series and event sequences without data transformation, learning, or ad hoc discretization. Our method models the continuous-discrete duality of real-world time series to handle quantization and repeated-value artifacts and introduces a latent event clustering strategy to mitigate bias from event co-occurrence and redundancy. Together, these yield a robust and unified framework that bridges discrete and continuous mutual information. We evaluate the proposed estimator on four representative tasks: discrete-continuous time-delayed mutual information for causality analysis, global and local temporal repetition discovery, discrete covariate selection for time series forecasting, and continuous feature selection for classification. Experiments on synthetic and real-world datasets show consistent improvements over existing methods in accuracy, robustness, and interpretability, positioning our approach as a general-purpose dependence operator for heterogeneous temporal data, similar to Pearson correlation for homogeneous time series. Code available at: https://github.com/HaojiHu/Multimodal-Temporal-Data-Quantification

---

### EIVE: End-to-End Instance-Specific Visual Explanations for Detection Transformers

Authors: Jianlin Xiang, Yanshan Li, Linhui Dai
Published: 2026-06-01
Categories: cs.CV
arXiv: https://arxiv.org/abs/2606.01601v1

Visual explainability for object detection remains challenging due to the multi-instance nature of detection. Existing approaches predominantly adopt post-hoc paradigms, such as gradient-based or perturbation-based explanation methods, to interpret pretrained detectors. However, these methods require additional gradient computation or repeated model inference, resulting in limited efficiency. To address this issue, we propose an End-to-end Instance-specific Visual Explanation framework (EIVE) that directly generates instance-level saliency maps following the forward pass of Detection Transformer (DETR)-like models. Specifically, we reformulate the cross-attention mechanism in the decoder as an instance-level feature attribution pathway, so that the cross-attention of each object query corresponds to the visual attribution of its predicted instance. Based on this formulation, we design a cross-layer hybrid consensus fusion (CLHCF) module to aggregate cross-attention signals across decoder layers, producing stable and compact explanations. The explanation process of EIVE requires neither gradient computation nor input perturbation, yielding high computational efficiency, and applies to single- and multi-scale DETR-like object detectors. Finally, we present an attention-aware joint training strategy (AAJTS) as a training-oriented application, which imposes spatial constraints on cross-attention patterns to encourage stable and concentrated attribution representations, thereby improving both interpretability and detection performance. Experiments on MS COCO 2017, ExDark, and Cityscapes demonstrate that EIVE produces high-quality instance-level saliency maps and achieves performance comparable to, or better than, state-of-the-art post-hoc methods across standard metrics, while substantially improving explanation efficiency. Code is available at https://github.com/xjlDestiny/EIVE.git.

---

### RoboTrustBench: Benchmarking the Trustworthiness of Video World Models for Robotic Manipulation

Authors: Huiqiong Li, Jiayu Wang, Zhiting Mei, Anirudha Majumdar, Jingjing Chen, Bin Zhu
Published: 2026-06-01
Categories: cs.CV, cs.CL, cs.RO
arXiv: https://arxiv.org/abs/2606.01600v2

Video world models are increasingly used in robotic manipulation, yet existing benchmarks mostly evaluate them under valid, feasible, and safe instructions. We introduce RoboTrustBench, a benchmark for evaluating the trustworthiness of video world models under four scenarios: Normal, Constraint-Sensitive, Counterfactual, and Adversarial. Built from real-world DROID episodes, RoboTrustBench contains 1,207 expert-validated instruction-image pairs and a six-dimensional evaluation protocol with 13 fine-grained criteria. Evaluating seven representative video world models with human and MLLM assessment, we find that current models often generate visually coherent videos, but struggle with constraint reasoning, counterfactual grounding, physical interaction, and unsafe-instruction suppression. These results show that visual quality and surface-level instruction following are insufficient for trustworthy robotic video world modeling.

---

### RoboTrustBench: Benchmarking the Trustworthiness of Video World Models for Robotic Manipulation

Authors: Huiqiong Li, Jiayu Wang, Zhiting Mei, Anirudha Majumdar, Jingjing Chen, Bin Zhu
Published: 2026-06-01
Categories: cs.CV, cs.CL, cs.RO
arXiv: https://arxiv.org/abs/2606.01600v1

Video world models are increasingly used in robotic manipulation, yet existing benchmarks mostly evaluate them under valid, feasible, and safe instructions. We introduce RoboTrustBench, a benchmark for evaluating the trustworthiness of video world models under four scenarios: Normal, Constraint-Sensitive, Counterfactual, and Adversarial. Built from real-world DROID episodes, RoboTrustBench contains 1,207 expert-validated instruction-image pairs and a six-dimensional evaluation protocol with 13 fine-grained criteria. Evaluating seven representative video world models with human and MLLM assessment, we find that current models often generate visually coherent videos, but struggle with constraint reasoning, counterfactual grounding, physical interaction, and unsafe-instruction suppression. These results show that visual quality and surface-level instruction following are insufficient for trustworthy robotic video world modeling.

---

### TRON: Targeted Rule-Verifiable Online Environments for Visual Reasoning RL

Authors: Tianze Yang, Yucheng Shi, Ruitong Sun, Jingyuan Huang, Ninghao Liu, Jin Sun
Published: 2026-06-01
Categories: cs.AI
arXiv: https://arxiv.org/abs/2606.01599v1

Reinforcement learning (RL) for visual reasoning needs scalable, verifiable, and controllable training signals. Existing visual RL post-training trains on static curated datasets, with fixed image-question-answer samples bounded by their collection budget. In this work, we introduce TRON (Targeted, Rule-verifiable Online eNvironments), an online environment substrate: a training rollout is generated on demand by a controllable generator-verifier program that samples a fresh latent visual state, renders an image, asks a question, and exactly verifies the answer. A single run can therefore draw an unbounded stream of fresh instances at the difficulty level required by the current curriculum. The current TRON suite contains 520 environments organized into five ability buckets (spatial, mathematical, diagram, pattern/logic, and counting); the same substrate supports both a single full model trained on all buckets and per-bucket ability-specialist models, with no additional data collection. We also introduce a substrate analysis covering generation reliability, instance and level diversity, cross-environment near-duplicates, and base-model pass rate by difficulty level. RL post-training with METHOD consistently improves performance on ten external multimodal reasoning benchmarks across Qwen3-VL-4B, Qwen2.5-VL-7B, and MiMo-VL-7B-SFT.

---