×

Academic Intelligence · Curated Daily

Explore the Frontier of Global Academia

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

Authors: Bo Yan ×
Shuffle
01.
arXiv (CS.CL) 2026-06-19

Where Does Social Reasoning Come From? Capability Provenance in Language Models

We use training-data attribution as an interpretable tool for capability discovery, mapping which regions of the pretraining corpus support social-reasoning versus STEM-reasoning in OLMo3-7B. Training-data attribution measures how strongly each training document influences a model's predictions on a benchmark, but document-level scores are too noisy to identify which corpus regions support which capabilities, and prior work has emphasized factual knowledge rather than reasoning. We compute gradient-based attribution (TrackStar via Bergson) over a working set drawn from the de-duplicated Dolma3 mix, aggregate influence across WebOrganizer's 24-format x 24-topic taxonomy (576 bins), and contrast benchmark pairs in a 2x2 design that varies domain (social vs. STEM) and capability type (reasoning vs. knowledge): SocialIQA and MMLU Social Sciences against ARC-Challenge and MMLU STEM. Social and STEM reasoning draw on qualitatively distinct corpus regions, and the contrast is sharper at the reasoning level than at the knowledge level. Targeted machine unlearning provides partial causal validation: forgetting high-attribution topic bins (e.g., Literature for SocialIQA) degrades the aligned benchmark more than within-bin random baselines, and we open-source all code, sampling manifests, the bin-level influence matrix, and unlearning checkpoints.

02.
arXiv (CS.CL) 2026-06-16

Enhancing LLM Safety Through a Theoretical Minimax Game Lens

The rapid advancement of large language models (LLMs) necessitates effective mechanisms to ensure their responsible deployment by accurately distinguishing unsafe content from benign content. While substantial safety datasets are available in English, multilingual safety modeling remains underexplored due to limited open-source safety datasets in other languages. Even within English datasets, safe yet sensitive corner-case content is scarce, leading to shortcut learning by models and non-trivial false-positive rates. To mitigate these issues, we introduce a novel minimax reinforcement learning (RL) framework wherein a data generator and a classifier model co-evolve, facilitating the production of high-quality synthetic multilingual safety data. We theoretically formalize this interaction as a minimax game and rigorously demonstrate convergence to a Nash equilibrium. Empirical evaluations confirm that our synthetic data generation method significantly enhances the classifier model performance, enabling a substantially smaller model to surpass the state-of-the-art by nearly 10% on English benchmarks while achieving 4.5x faster inference speed. These results establish a scalable and efficient methodology for synthetic data generation, advancing the development of safer and more robust multilingual LLM deployments.

03.
arXiv (CS.AI) 2026-06-17

From Brewing to Resolution: Tracing the Internal Lifecycle of Code Reasoning in LLMs

arXiv:2606.17648v1 Announce Type: new Abstract: Standard accuracy metrics cannot explain why LLMs handle variable tracking but fail on semantically equivalent loops. We study an internal lifecycle of code reasoning in which models first brew the answer, making it linearly recoverable many layers before it becomes self-decodable, and then diverge into one of four resolution outcomes: Resolved, Overprocessed, Misresolved, or Unresolved. Understanding this lifecycle matters because similar task accuracies can mask fundamentally different failure modes that surface-level evaluation cannot detect. We introduce a dual diagnostic framework pairing layer-wise linear probing with Context-Stripped Decoding (CSD) and apply it to six code-reasoning task families across 16 models spanning Qwen, Llama, and DeepSeek architectures. All four outcomes carry substantial mass in every task family: overall Resolved is only 41.5%, with multiple tasks below 30%. Controlled sweeps over structure, depth, and operators expose task-specific failure bottlenecks: Function Call Resolved plunges from 61.1% to 2.5% as call depth increases from one to three. Across architectures and scales, the brewing scaffold remains stable, with normalized brewing duration 24-42% across all 16 models, while resolution success varies with capability. This indicates that the scaffold is a stable empirical regularity across the tested decoder-only Transformer families, whereas resolution success covaries with capability, scale, and training. Code: https://github.com/euyis1019/llm-brewing

04.
arXiv (CS.CL) 2026-06-15

TVIR: Building Deep Research Agents Towards Text-Visual Interleaved Report Generation

Deep Research Agents have shown strong capability in multi-step information retrieval, reasoning, and long-form report generation, but existing benchmarks and systems remain predominantly text-centric, with limited evaluation of whether visual elements are factually reliable and well aligned with the surrounding analysis. To address this gap, we introduce TVIR (Text-Visual Interleaved Report Generation), which includes TVIR-Bench, a benchmark of 100 expert-curated multimodal deep research tasks that require visual elements to serve specific analytical sub-goals, and TVIR-Agent, a hierarchical multi-agent framework that serves as a strong baseline for constructing outlines, retrieving images, generating charts with traceable sources, and composing reports through context-aware sequential writing. We further develop a dual-path evaluation framework that combines Textual Assessment and Visual Assessment. Experiments across nine deep research systems show that TVIR-Agent achieves strong overall performance, underscoring the importance of explicit multimodal design and evaluation for evidence-driven report generation.

05.
arXiv (CS.CV) 2026-06-17

Gaussian Light Field Splatting: A Physical Prior-Driven Vision Transformer for Unsupervised Low-Light Image Enhancement

Existing unsupervised low-light image enhancement methods often encounter local exposure imbalance and color distortion under complex non-uniform illumination. In addition, most Vision Transformers lack an explicit mechanism for modeling the physical priors of illumination degradation. To address these limitations, we propose GLFS, a Gaussian light field splatting-based Vision Transformer that integrates continuous physical illumination modeling from Gaussian splatting into the Transformer architecture. In GLFS, scene illumination is represented by a superposition of anisotropic Gaussian basis functions. Physics-guided biases are introduced into self-attention to adaptively infer a spatial gain field, enabling accurate and uniform restoration under complex illumination. To reduce color bias and structural degradation during enhancement, a color-vector angular loss and a luminance-edge loss are further developed. These losses enforce hue consistency and improve the structural fidelity of local details. Extensive ablation studies and quantitative evaluations show that GLFS provides clear advantages in illumination correction and detail preservation. It achieves state-of-the-art performance and offers a new representation paradigm for low-light image enhancement.

06.
arXiv (CS.CV) 2026-06-19

OncoReg: Medical Image Registration for Oncological Challenges

In modern cancer research, the vast volume of medical data generated is often underutilised due to challenges related to patient privacy. The OncoReg Challenge addresses this issue by enabling researchers to develop and validate image registration methods through a two-phase framework that ensures patient privacy while fostering the development of more generalisable AI models. Phase one involves working with a publicly available dataset, while phase two focuses on training models on a private dataset within secure hospital networks. OncoReg builds upon the foundation established by the Learn2Reg Challenge by incorporating the registration of interventional cone-beam computed tomography with standard planning fan-beam CT images in radiotherapy. Accurate image registration is crucial in oncology, particularly for dynamic treatment adjustments in image-guided radiotherapy, where precise alignment is necessary to minimise radiation exposure to healthy tissues while effectively targeting tumours. This work details the methodology and data behind the OncoReg Challenge and provides a comprehensive analysis of the competition entries and results. Findings reveal that feature extraction plays a pivotal role in this registration task. A new method emerging from this challenge demonstrated its versatility, while established approaches continue to perform comparably to newer techniques. Both deep learning and classical approaches still play significant roles in image registration, with the combination of methods, particularly in feature extraction, proving most effective.

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

CausalMoE: A Billion-Scale Multimodal Foundation Model for Granger Causal Discovery with Pattern-Routed Heterogeneous Experts

arXiv:2606.13024v1 Announce Type: cross Abstract: Granger Causal Discovery (GCD) is fundamental for analyzing temporal dependencies in complex systems. However, existing neural GCD methods predominantly rely on a "one-size-fits-all" paradigm, struggling to capture distribution shifts and dynamic regime changes inherent in real-world time series. This often leads to entangled representations and spurious causal graphs. In this paper, we propose CausalMoE, a billion-scale multimodal Granger causal foundation model that explicitly models patch-level heterogeneity. CausalMoE introduces a Pattern-Routed Mixture of Heterogeneous Experts, which dynamically identifies latent temporal patterns and routes patches to specialized domain experts, effectively decoupling regime-specific mechanisms from shared dynamics. To ensure interpretable graph recovery, we design a Causality-Aware Self-Attention mechanism operating across variables, yielding sparse Granger causal graphs via proximal optimization. Furthermore, CausalMoE is the first to integrate LLMs and VLMs to align numerical signals with textual and visual priors, regularizing causal estimation in complex scenarios. Extensive experiments demonstrate that CausalMoE establishes a new state-of-the-art on fully supervised benchmarks, while effectively generalizing to few-shot settings where traditional methods fail.

08.
arXiv (CS.CV) 2026-06-16

MotionVLA: Vision-Language-Action Model for Humanoid Motion

Generating realistic humanoid motion from scene images and text involves both low-frequency pose semantics and high-frequency physical dynamics. However, many existing methods tokenize motion with a single shared codebook, forcing heterogeneous motion signals into the same quantization space. Our frequency-domain analysis of human motion data reveals a clear mismatch between single-codebook quantization and motion statistics: five DCT coefficients capture 93% of joint-position energy but only 37% of joint-velocity energy, which can bias quantization toward pose statistics and under-represent high-frequency velocity components. A second challenge lies in adapting a standard autoregressive model to effectively model high-frequency physical signals in motion sequences. Therefore, we propose DSFT, a dual-stream frequency tokenizer that separates motion into Base and physical streams and compresses them independently with DCT truncation and BPE. Furthermore, we present MotionVLA, a Qwen3.5-based model that arranges Base and physical tokens in a unified sequence, where Phys tokens are predicted after Base tokens. Experiments on HumanML3D and MBench show that, despite using a lightweight 2B backbone, MotionVLA reduces the Diversity gap to real data by over 50% on HumanML3D and improves Motion-Condition Consistency by 3.8% on MBench, supporting frequency-aware dual-stream decoupling as an effective formulation for autoregressive motion generation. Code: https://github.com/AIGeeksGroup/MotionVLA. Website: https://aigeeksgroup.github.io/MotionVLA.

09.
arXiv (CS.AI) 2026-06-12

MARS: Margin-Adversarial Risk-controlled Stopping for Parallel LLM Test-time Scaling

arXiv:2606.12935v1 Announce Type: new Abstract: Parallel test-time scaling samples many reasoning traces and majority-votes their answers, improving LLM accuracy but requiring traces to run to completion, incurring substantial computational overhead. We observe that probing partial traces at intermediate checkpoints can extract current answers without disrupting generation, revealing an evolving aggregate vote. Based on this observation, we introduce MARS, a margin-adversarial stopping rule that estimates which active traces are likely to change their answers and stops once the leader remains safe under a conservative bound on future vote movement. The rule separates two sources of uncertainty. It learns the trace-level switch probabilities that determine how much of the current margin is likely to be retained, while handling the harder question of where switching traces land through an adversarial bound calibrated from warmup traces. With true switch probabilities, MARS guarantees with high probability that the early-stopped answer matches the full-budget vote. In practice, a five-feature logistic model closely matches oracle switching behavior. Across three reasoning models and three competition-math benchmarks, MARS saves 25-47% of self-consistency tokens and 14-29% on top of DeepConf Online, a strong confidence-weighted baseline that already filters and truncates weak traces, while matching the accuracy of the corresponding full-budget baselines.

10.
arXiv (CS.AI) 2026-06-16

Frontier: Towards Comprehensive and Accurate LLM Inference Simulation

arXiv:2605.21312v2 Announce Type: replace-cross Abstract: Modern LLM serving is no longer homogeneous or monolithic. Production systems now combine disaggregated execution, complex parallelism, runtime optimizations, and stateful workloads such as reasoning, agents, and RL rollouts. Simulation is attractive for exploring this growing design space, yet existing simulators lack the architectural completeness and decision-grade fidelity it demands. Their monolithic-replica abstractions are ill-suited to disaggregated serving, while average-case analytical proxies can distort SLA predictions and even reverse optimization conclusions. We present Frontier, a discrete-event simulator for modern LLM inference serving. Frontier features a disaggregated abstraction. It captures the structure and dynamics of modern serving systems by modeling co-location, Prefill-Decode Disaggregation (PDD), and Attention-FFN Disaggregation (AFD) with role-specific cluster workers, incorporating key runtime optimizations (e.g., CUDA Graphs, speculative decoding) within the scheduler-batch-engine loop, and supporting stateful requests for emerging workloads. It further provides accurate and generalizable predictions of computation, communication, and memory costs across diverse serving scenarios with complex workload compositions. On 16-H800 GPU testbed, Frontier achieves an average throughput error below 4%. Compared with state-of-the-art simulators, it reduces end-to-end latency error from 44.9% to 6.4% under co-location and from 51.7% to 2.6% under disaggregation. It scales to over 1K GPUs on commodity CPUs and enables new use cases such as SLA-dependent Pareto frontier exploration, heterogeneous disaggregated allocation, agentic reasoning scheduling validation, and RL post-training reconfiguration. We release Frontier at https://github.com/NetX-lab/Frontier.

11.
arXiv (CS.AI) 2026-06-18

HeRo-Q: A General Framework for Stable Low Bit Quantization via Hessian Conditioning

arXiv:2601.21626v2 Announce Type: replace-cross Abstract: Post Training Quantization (PTQ), a mainstream model compression technique, often leads to the paradoxical 'low error, high loss' phenomenon because it focuses solely on minimizing quantization error. The root cause lies in the Hessian matrix of the LLM loss landscape: a few high curvature directions are extremely sensitive to perturbations. To address this, we propose the Hessian Robust Quantization (HeRo Q) algorithm, which applies a lightweight, learnable rotation-compression matrix to the weight space prior to quantization. This joint framework reshapes the loss landscape by reducing the largest Hessian eigenvalue and reducing its max eigenvalue, thereby significantly enhancing robustness to quantization noise. HeRo-Q requires no architectural modifications, incurs negligible computational overhead, and integrates seamlessly into existing PTQ pipelines. Experiments on Llama and Qwen models show that HeRo Q consistently outperforms state of the art methods including GPTQ, AWQ, and SpinQuant not only achieving superior performance under standard W4A8 settings, but also excelling in the highly challenging W3A16 ultra low bit regime, where it boosts GSM8K accuracy on Llama3 8B to 70.15\% and effectively avoids the logical collapse commonly seen in aggressive quantization.

12.
medRxiv (Medicine) 2026-06-15

International Consensus Guideline on Management of Genitourinary Adverse Events Associated with Prostate Cancer Radiotherapy

Purpose/Objective: Genitourinary (GU) adverse events (AEs) are common during and after pelvic radiation therapy (RT) for prostate cancer and can substantially impact quality of life. We convened an international committee to establish consensus in the prevention, mitigation, and management of radiation-related acute and late GU AEs, as there are no relevant evidence-based consensus guidelines to inform treating providers. Materials/Methods: A systematic evidence review focused on mitigation and management of radiation-related acute and late GU AEs was performed in PubMed, Embase and Cochrane. The following topics were addressed: management of acute GU AEs in the intact and post-operative settings; RT techniques; bladder outlet obstruction procedures; and indications for urology referral or hyperbaric oxygen therapy (HBO). Evidence-based consensus recommendations were developed using a Delphi process. We highlight the current state of evidence and evidence gaps worthy of future study. Results: Consensus was reached for 31 key questions. For management of lower urinary tract symptoms (LUTS), most evidence comes from trials in patients without cancer and not undergoing RT. A consensus algorithm for medical management of acute GU AEs was developed with the following highlights: (a) alpha blockers as 1st-line for obstructive symptoms in the intact setting, (b) anti-spasmodics as 1st -line for irritative symptoms in the intact setting, and (c) anti-spasmodics as 1st -line in the post-operative setting. The consensus algorithm provides an ordered list of medications to offer if 1st -line options afford inadequate relief. For RT fractionation, randomized clinical trial (RCT) data are available. 40% of panelists rarely or never use standard fractionation over moderate hypofractionation for patients with baseline LUTS, but most consider moderate hypofractionation over SBRT for AUA IPSS > 15. For patients with severe obstructive LUTS (most commonly AUA IPSS >20), the panel recommends a prophylactic bladder outlet obstruction procedure and, if obstructive symptoms improve, consideration of moderate hypofractionation or SBRT, based on retrospective data. There is one RCT supporting use of HBO for late radiation cystitis. Conclusions: The consensus guideline synthesizes available evidence and expert opinion across key clinical decision points to provide practical guidance in the prevention, mitigation, and management of radiation-related acute and late GU AEs in prostate cancer RT. Envisioned as a living document with periodic updates, this guideline serves as a resource for practicing radiation oncologists by outlining expert-derived consensus recommendations of evidence-based care in areas where high-quality data is limited.

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

DataMagic: Transforming Tabular Data into Data Insight Video

arXiv:2606.20388v1 Announce Type: cross Abstract: Data videos integrate dynamic charts, voice narration, and synchronized animations to communicate data insights as temporal narratives, making them an effective medium for improving data consumption efficiency in the data management lifecycle. However, producing high-quality data videos requires expertise spanning data analysis, narrative design, and video production. Existing approaches fall short: static visualization tools (e.g., BI dashboards) lack narrative logic and animation; authoring tools require users to pre-prepare visualizations rather than working from raw data; pixel-level video generation models cannot guarantee data fidelity or provenance. We demonstrate DataMagic, an end-to-end interactive system that transforms raw tabular data and natural language queries into narrative data-insight videos. To ensure data fidelity, DataMagic introduces the declarative specification DVSpec, which binds visual and animation elements to underlying data fields through data-driven semantic references. To address the combinatorial explosion of the design space, DataMagic adopts a Generate-then-Orchestrate multi-agent architecture that generates candidate scenes in parallel and then optimizes narrative coherence through global orchestration. Leveraging DVSpec's decoupling of logic and rendering, the system further supports three interaction modes and structured provenance-based data Q&A, transforming one-way videos into explorable interactive data interfaces. Evaluation on 109 real-world samples validates the effectiveness of the DataMagic. Homepage: https://datamagic-home.github.io/

14.
arXiv (CS.AI) 2026-06-19

TerraMind: Large-Scale Generative Multimodality for Earth Observation

arXiv:2504.11171v5 Announce Type: replace-cross Abstract: We present TerraMind, the first any-to-any generative, multimodal foundation model for Earth observation (EO). Unlike other multimodal models, TerraMind is pretrained on dual-scale representations combining both token-level and pixel-level data across modalities. On a token level, TerraMind encodes high-level contextual information to learn cross-modal relationships, while on a pixel level, TerraMind leverages fine-grained representations to capture critical spatial nuances. We pretrained TerraMind on nine geospatial modalities of a global, large-scale dataset. In this paper, we demonstrate that (i) TerraMind's dual-scale early fusion approach unlocks a range of zero-shot and few-shot applications for Earth observation, (ii) TerraMind introduces "Thinking-in-Modalities" (TiM) – the capability of generating additional artificial data during finetuning and inference to improve the model output – and (iii) TerraMind achieves beyond state-of-the-art performance in community-standard benchmarks for EO like PANGAEA. The pretraining dataset, the model weights, and our code are open-sourced under a permissive license.

15.
arXiv (CS.CV) 2026-06-17

Seeing Is Not Screening: Multimodal Hidden Instruction Attacks on Agent Skill Scanners

Agent skills are emerging as an important attack surface in LLM-based systems. Through an empirical study of existing skill scanners, we find that current defenses primarily rely on textual descriptions, manifests, and source code as the main signals for security analysis, which can leave visually conveyed malicious intent insufficiently examined. This creates a practical blind spot: harmful operational instructions hidden in images may bypass scanning while still being recoverable by multimodal agents during deployment. To systematically investigate this threat, we propose SkillCamo, a document-mediated multimodal instruction attack that conceals malicious instructions within images bundled with a skill while rewriting the surrounding documentation to naturally reference those images as part of the normal workflow. Thus, the attack does not rely on the image alone, but on the joint interpretation of textual guidance and visual payload at execution time. To defend against such attacks, we further propose ExecScan, an execution-grounded multimodal scanning module that performs intent extraction, behavior reconstruction, abuse assessment, and deliberative execution simulation over skill artifacts. ExecScan jointly analyzes documentation, code, referenced resources, and visual content to recover hidden instructions, reconstruct executable behavior chains, and identify downstream risks such as exfiltration, destruction, persistence, deception, and privilege escalation. Extensive experiments show that image-hidden malicious instructions challenge existing skill scanners, while ExecScan can improve the skill scanning performance.

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

A Context-Aware Dataset for Stance Detection in Bioethical Controversies on Reddit

Bioethical debates increasingly unfold on social media, yet stance detection research lacks large-scale, domain-specific resources for modeling such context-dependent discourse. We present BioStance, a context-aware dataset of 39,600 annotated Post-Comment pairs from Reddit bioethical discussions. BioStance covers six controversial targets across three dimensions of bioethical controversy: fundamental value conflicts, individual liberty versus collective responsibility, and technological uncertainty. Each instance preserves hierarchical conversational context and is labeled by three independent annotators using a three-class stance scheme: Favor, Against, and None. The annotations achieve a mean Krippendorff's $\alpha$ of 0.82, indicating substantial reliability. By combining thematic diversity, conversational structure, and high-quality human annotation, BioStance supports research on context-aware stance detection, argument mining, and computational analysis of bioethical discourse.

17.
arXiv (CS.CV) 2026-06-16

Prompt Disentanglement via Language Guidance and Representation Alignment for Domain Generalization

Domain Generalization (DG) seeks to develop a versatile model capable of performing effectively on unseen target domains. Notably, recent advances in pre-trained Visual Foundation Models (VFMs), such as CLIP, have demonstrated considerable potential in enhancing the generalization capabilities of deep learning models. Despite the increasing attention toward VFM-based domain prompt tuning within DG, the effective design of prompts capable of disentangling invariant features across diverse domains remains a critical challenge. In this paper, we propose addressing this challenge by leveraging the controllable and flexible language prompt of the VFM. Noting that the text modality of VFMs is naturally easier to disentangle, we introduce a novel framework for text feature-guided visual prompt tuning. This framework first automatically disentangles the text prompt using a large language model (LLM) and then learns domain-invariant visual representation guided by the disentangled text feature. However, relying solely on language to guide visual feature disentanglement has limitations, as visual features can sometimes be too complex or nuanced to be fully captured by descriptive text. To address this, we introduce Worst Explicit Representation Alignment (WERA), which extends text-guided visual prompts by incorporating an additional set of abstract prompts. These prompts enhance source domain diversity through stylized image augmentations, while alignment constraints ensure that visual representations remain consistent across both the original and augmented distributions. Experiments conducted on major DG datasets, including PACS, VLCS, OfficeHome, DomainNet, and TerraInc, demonstrate that our proposed method outperforms state-of-the-art DG methods.

18.
arXiv (CS.CL) 2026-06-11

GrowLoop: Self-Evolving Conversation Evaluation Seeded by Human

With the rapid advancement of large language models, evaluating human-likeness in open-ended conversation has become increasingly important. However, human-likeness is a form of tacit knowledge that humans perceive intuitively, yet the underlying criteria resist explicit formulation. Human judgments vary widely, with strong agreement on some cases and legitimate disagreement on others. Meanwhile, the criteria behind human judgments remain implicit, leaving no clear basis for constructing cases. Further, what counts as human-likeness is not static, but evolving with model capability and human expectations. Despite progress in evaluation methods such as expert-authored benchmarks, Reward Models, and self-evolving benchmarks, none addresses all three challenges simultaneously. Therefore, we propose GrowLoop, a self-evolving conversation evaluation system that continuously adapts as models advance and scenarios shift. Starting from minimal human seed annotations, LLM agents iteratively extract and refine evaluation rubrics through Heuristic Learning. Human-AI agreement is required where annotators converge, while only plausibility is expected where they diverge. Moreover, the Rubric-Case co-evolution mechanism enables continuous evolution. When the evaluation target shifts, new human seeds expand the system's coverage accordingly. When applied to human-likeness evaluation in open-ended conversation, the AI judge guided by these rubrics not only substantially outperforms existing methods in alignment with human judgments, but also uncovers issues that annotators overlook. The resulting benchmark effectively discriminates models across capability tiers and reveals where they fall short, while generalizing to new scenarios and adapting as models advance. Our work shifts the benchmarking paradigm from manual updates or difficulty scaling to comprehensive, continuous self-evolution.

19.
arXiv (CS.AI) 2026-06-11

nD-RoPE: A Generalized RoPE for n-Dimensional Position Embedding

arXiv:2606.12146v1 Announce Type: cross Abstract: Rotary Position Embedding (RoPE) is widely adopted in Transformer models, yet its extension to high-dimensional domains lacks a unified theoretical formulation. Most existing approaches either apply rotations independently along each axis or empirically mix frequencies, which limits cross-dimensional interactions and yields direction-dependent representations. To address these limitations, we propose nD-RoPE, a decomposition-free generalization of RoPE to arbitrary dimensions. From a translation-invariant formulation in continuous Hilbert space, we derive a spectral condition for isotropy that requires treating positions and frequencies as coupled \(n\)-dimensional vectors. We instantiate this formulation with a multi-scale regular-simplex wave-vector design, which provides non-degenerate spatial coverage and a symmetric, directionally balanced second-order response. Experiments across images, videos, and point clouds demonstrate consistent performance gains and improved generalization in high-dimensional settings.

20.
arXiv (CS.CL) 2026-06-15

AgentSpec: Understanding Embodied Agent Scaffolds Through Controlled Composition

LLM agents are increasingly built not as single model calls, but as scaffolded systems that combine reasoning, memory, reflection, action execution, and learning. While such scaffolds often improve performance, they are often embedded in tightly coupled pipelines, making it difficult to isolate component contributions, compare alternative designs, or understand how module interactions shape agent behavior. We introduce AgentSpec, a modular specification framework that represents embodied agents as typed compositions of reusable policy components with standardized interfaces. AgentSpec standardizes the interfaces among perception, memory, reasoning, reflection, action, and optional learning, enabling components to be swapped and recombined under controlled conditions. We instantiate this framework across DeliveryBench, ALFRED, MiniGrid, and RoboTHOR, and analyze reasoning, memory, reflection, and reinforcement-learning modules across model backbones. Our results show that agent performance is governed by scaffold compatibility and interaction effects rather than isolated module strength. In particular, structured multi-granularity memory improves long-horizon state tracking, reasoning and memory interact non-uniformly across environments, reflection trades off correction and cost, and RL-trained policies compose best when optimized with deployment-time scaffold structure. AgentSpec provides a controlled foundation for studying, comparing, and designing composable LLM agents. Our code, baselines and interactive playground are publicly available at https://agentspec-embodied.github.io.

21.
arXiv (CS.CL) 2026-06-11

Breaking Entropy Bounds: Accelerating RL Training via MTP with Rejection Sampling

Reinforcement learning (RL) has become a key component in modern large language models, yet the rollout stage remains the key bottleneck in RL training pipelines. Although Multi-Token Prediction (MTP) offers a natural solution to accelerate rollouts through speculative decoding, many studies have observed that MTP acceptance rates degrade significantly during RL training, leading to limited speedup performance. To address this bottleneck, we present Bebop, a systematic study of MTP in LLM post-training, and offer practical recipes to integrate MTP into large-scale RL pipelines. First, we reveal that the MTP acceptance rate is fundamentally bounded by the fluctuation of model entropy, which demonstrates a clear negative linear relationship with the rise of entropy in the RL stage. Second, we show that probabilistic rejection sampling largely alleviates the disturbance introduced by entropy in RL compared to greedy draft sampling. We further identify that the conventional MTP training objectives (cross-entropy or KL) are suboptimal in such settings, and therefore we propose a novel end-to-end TV loss that directly optimizes multi-step rejection sampling acceptance rate, yielding ~10% acceptance rate improvements, achieving up to 95% acceptance rates and up to 25% extra inference throughput gains across mathematical reasoning, code generation, and agentic tasks. Third, we test various online MTP training strategies during RL and show that pre-RL MTP training with e2e TV loss and rejection sampling achieves a consistent acceptance rate and speedup throughout the entire RL, eliminating the need for costly online MTP updating. We provide extensive experiments and analysis that validate our findings. Experimental results show our method achieves up to 1.8x end-to-end acceleration in async RL training of Qwen3.5, Qwen3.6, and Qwen3.7 models.

22.
arXiv (CS.CL) 2026-06-19

DeepSeek-V4: Towards Highly Efficient Million-Token Context Intelligence

We present a preview version of DeepSeek-V4 series, including two strong Mixture-of-Experts (MoE) language models – DeepSeek-V4-Pro with 1.6T parameters (49B activated) and DeepSeek-V4-Flash with 284B parameters (13B activated) – both supporting a context length of one million tokens. DeepSeek-V4 series incorporate several key upgrades in architecture and optimization: (1) a hybrid attention architecture that combines Compressed Sparse Attention (CSA) and Heavily Compressed Attention (HCA) to improve long-context efficiency; (2) Manifold-Constrained Hyper-Connections (mHC) that enhance conventional residual connections; (3) and the Muon optimizer for faster convergence and greater training stability. We pre-train both models on more than 32T diverse and high-quality tokens, followed by a comprehensive post-training pipeline that unlocks and further enhances their capabilities. DeepSeek-V4-Pro-Max, the maximum reasoning effort mode of DeepSeek-V4-Pro, redefines the state-of-the-art for open models, outperforming its predecessors in core tasks. Meanwhile, DeepSeek-V4 series are highly efficient in long-context scenarios. In the one-million-token context setting, DeepSeek-V4-Pro requires only 27% of single-token inference FLOPs and 10% of KV cache compared with DeepSeek-V3.2. This enables us to routinely support one-million-token contexts, thereby making long-horizon tasks and further test-time scaling more feasible. The model checkpoints are available at https://huggingface.co/collections/deepseek-ai/deepseek-v4.

23.
arXiv (CS.CV) 2026-06-17

ActWorld: From Explorable to Interactive World Model via Action-Aware Memory

Interactive world models aim to simulate environment dynamics under real-time user actions. However, their action vocabulary is largely confined to navigation: most actions correspond to motion (e.g., walk, turn, look around), while interaction with objects in the scene (e.g., pick up plates, open doors, or trigger physical responses) is either absent, restricted to game domains, or relegated to prompt-to-full-video scenarios. The resulting worlds are visually explorable but not truly actionable. In this work, we present ActWorld, an interactive world model that extends prior navigation-centric generators to support mid-rollout object interaction within a chunk-autoregressive framework. We argue that the navigation-interaction gap stems from two bottlenecks. First, a data bottleneck: the lack of human-object interaction data with accurate, dense labels. Second, a memory bottleneck: recency-biased history compression in existing world models discards the event-transition frames that causally determine subsequent object states, leading to an action-forgetting pathology. On the data side, we construct a 100K interaction video dataset, each annotated with per-chunk captions via chain-of-thought reasoning. On the model side, we introduce a hierarchical action-aware memory design that routes history compression by interaction importance, complemented by a persistent memory bank that maintains event-update and object-identity tokens across long rollouts. Experiments show that ActWorld supports both flexible navigation and rich object interaction within a single model, substantially improving interaction fidelity over navigation-only baselines without sacrificing viewpoint control. Project page is available at https://interactwm.github.io/ActWorld.

24.
arXiv (CS.LG) 2026-06-16

Scalable Graph Condensation with Evolving Capabilities

arXiv:2502.17614v3 Announce Type: replace Abstract: The rapid growth of graph data creates significant scalability challenges as most graph algorithms scale quadratically with size. To mitigate these issues, Graph Condensation (GC) methods have been proposed to learn a small graph from a larger one, accelerating downstream tasks. However, existing approaches critically assume a static training set, which conflicts with the inherently dynamic and evolving nature of real-world graph data. This work introduces a novel framework for continual graph condensation, enabling efficient updates to the distilled graph that handle data streams without requiring costly retraining. This limitation leads to inefficiencies when condensing growing training sets. In this paper, we introduce GECC (\underline{G}raph \underline{E}volving \underline{C}lustering \underline{C}ondensation), a scalable graph condensation method designed to handle large-scale and evolving graph data. GECC employs a traceable and efficient approach by performing class-wise clustering on aggregated features. Furthermore, it can inherit previous condensation results as clustering centroids when the condensed graph expands, thereby attaining an evolving capability. This methodology is supported by robust theoretical foundations and demonstrates superior empirical performance. Comprehensive experiments including real world scenario show that GECC achieves better performance than most state-of-the-art graph condensation methods while delivering an around 1000$\times$ speedup on large datasets.

25.
arXiv (CS.AI) 2026-06-18

X+Slides: Benchmarking Audience-Conditioned Slide Generation

arXiv:2606.19256v1 Announce Type: new Abstract: Automatically generating slide decks from source documents is an important application of large language models (LLMs). Existing benchmarks primarily assess slide completeness and technical depth, while overlooking the target audience as a critical real-world factor. For instance, specialists demand rigorous proofs, whereas decision-makers prioritize actionable conclusions. To bridge this gap, we introduce X+Slides, a benchmark specifically designed for audience-conditioned slide generation. Built on a diverse corpus spanning 113 topics and seven presentation scenes, X+Slides employs a dynamic evaluation framework constructed from 8,133 deduplicated, source-grounded probes. By assigning audience-specific utility weights to the same source-grounded probes, X+Slides reports four complementary metrics: Audience Coverage measures how much audience-essential information is conveyed, Domain-wise Coverage shows which information types are covered, Efficiency measures delivered utility per unit of attention cost, and Correctness verifies whether slide claims are supported by the source. Experiments on DeepPresenter, SlideTailor, and NotebookLM show that current systems can recover a substantial but still incomplete part of audience-essential information: at $\tau_A=0.7$, DeepPresenter reaches a best Audience Coverage of 0.714, SlideTailor reaches 0.594, and the NotebookLM ablation reaches 0.853 while showing clear grounding differences. These results indicate that visual quality and broad topic coverage should not be treated as evidence support without source-grounded evaluation.