Open-AoE is a large-scale egocentric manipulation dataset with approximately 2,000 hours of video from over 500 contributors using 400+ smartphones, including detailed annotations and a toolchain for embodied learning.

This work introduces a multi-modal orchestration framework for semantic audio-driven humanoid control, enabling real-time autonomous selection of motion skills based on music or speech input. Validated on the Unitree G1 humanoid, it demonstrates robust sim-to-real transfer.

Researchers propose an adaptive control method for flexible joint robots with uncertain joint stiffness. The approach updates estimates of nonlinear torque-deflection relations using an implicit control law and a control-input-dependent regressor matrix, and analyzes robustness against motor position controller errors. Experiments on a flexible joint with nonlinear stiffness validate the approach.

MixCompress is a unified variable bit-rate (VBR) framework based on sparse structural specialization, combining sparsely gated Mixture-of-Experts (MoE) routing and Mixture-of-Depths (MoD) extension to dynamically scale model capacity, along with Conditional Auxiliary Transforms (CAT) for dynamic sub-band energy modulation. It addresses feature entanglement in existing VBR methods, achieves performance matching or surpassing single-rate baselines, and establishes a new Pareto frontier for computationally efficient image coding.

SD-MAR is a framework for training and evaluating vision-language models (VLMs) on multi-image analytical reasoning tasks. It constructs paired visual scenarios through controlled perturbations and generates reasoning tasks spanning semantic change attribution and quantitative comparison. Using GRPO-lite with Backward Discounted Allocation (BDA), a reinforcement learning approach that removes KL regularization, fine-tuning on SD-MAR improves in-domain accuracy by up to 36.95% on Qwen2.5-VL-7B and InternVL3-8B. Qwen2.5-VL-7B outperforms GPT-4.1 on the SD-MAR benchmark. Out-of-domain generalization is preserved or improved, with performance within 1% on MME, MMMU-Pro, MathVista and up to 4% improvement on MMBench. LLM-as-judge evaluation shows consistent improvements in logical coherence an…

This paper proposes DCVC-Mamba (DCVC-MB), a neural video codec framework for B-frame coding. It incorporates an IBP frame strategy for low-delay B-frame coding, a spatio-temporal fusion model based on state-space models for bidirectional temporal prediction, and an entropy-aware skipping mechanism that selectively omits coding certain latents to reduce entropy coding times. Two inference-time strategies are also implemented to enhance compression performance. Experimental evaluation shows that DCVC-MB achieves average BD-rate reductions of up to 8.98% compared to prior neural video codecs, and improvements of up to 30.45% and 1.81% over the VTM-19.0-LDP and VTM-19.0-RA (Inter-GoP=16) benchmarks, respectively, contributing to advances in neural video compression.

Addressing the challenge of defect segmentation in additive manufacturing XCT images, the proposed XCT-SAM framework sequentially adapts SAM using Conv-LoRA adapters, first on an alloy microstructure dataset then on XCT images, outperforming baselines on CycleGAN-XCT benchmarks and real NIST scans.

MonteRET is a region-aware retrieval-enhanced framework for generating chest CT findings sections. It integrates global and regional CT features, retrieves clinically relevant knowledge, and refines reports via a knowledge-guided rewriting agent. Evaluated on public and external cohorts, MonteRET improved report quality, semantic similarity, and clinical efficacy, with experts favoring its outputs.

Researchers introduce a new dataset and method for 3D lane detection in racing, leveraging multiple cameras and inertial odometry to achieve high-speed processing (300Hz) and improved accuracy, with F1 score >0.9 and reduced lateral errors.

This paper investigates whether vision foundation models build representations that reflect intrinsic properties of 3D Euclidean space. Instead of regressing depth or normals, the authors probe the relationship between visual feature space structure and Euclidean transformation group SE(3) using a mutual neighborhood metric and a Poincaré Adapter. They show that self-supervised vision models harbor latent subspaces strongly correlated with 3D space, even without 3D supervision. This leads to 'Latent-Space Navigation' techniques for visual odometry and localization without explicit 3D reconstruction.

The paper presents KeyFrame-Compass, the first comprehensive benchmark for evaluating keyframe-conditioned video generation, with 386 curated samples, an automated evaluation framework, and experiments revealing trade-offs between keyframe fidelity and naturalness.

This paper proposes SIRUS, a training-free inference-time framework for concept-level unlearning in text-to-video (T2V) models. SIRUS localizes target-related prompt evidence and suppresses target expression during sampling without updating the text encoder or denoising network. A video-oriented evaluation framework is introduced to separately measure target forgetting, non-target preservation, video quality, jailbreak robustness, and efficiency. On CogVideoX, SIRUS achieves 70.4% average forgetting success and 25.7% average frame hit, compared to 44.4%/47.2% for VideoEraser, while reducing the average VBench quality drop from -0.043 to -0.016. Transfer experiments on Wan2.2 suggest SIRUS generalizes across modern T2V backbones.

Multi-reference-to-audio-video (MR2AV) generation requires models to produce synchronized audio-video content conditioned on multiple references and textual instructions. Existing benchmarks focus on text-driven generation or single-reference preservation, lacking evaluation for MR2AV. This paper introduces MultiRef-Compass, a unified benchmark with 350 carefully curated samples covering multi-view subject preservation, multi-entity binding, and human-object-scene composition. It defines an evaluation protocol with four dimensions (Basic Quality, Reference Consistency, Audio-Visual Consistency, Instruction Following) and 14 sub-metrics, integrating automatic metrics with a rejudging-enhanced MLLM-as-a-Judge framework. Experiments on eight MR2AV systems reveal substantial room for improvem…

This paper introduces tool efficiency, a new quantitative metric to evaluate the rate of useful tool calls in an LLM agent trajectory. To ensure that tool efficiency is well-defined, it also introduces marginal tool utility, indicating per tool call whether it is useful or safely removable. The sign of marginal tool utility is determined using LLM-as-a-Judge. This work directly measures efficiency, complementing accuracy-based evaluations, and aims to inform future benchmark design and lean tool suite engineering.

Polestar is a training-free inference framework that addresses KV-cache reuse and decoding parallelism challenges in diffusion LLMs by leveraging token representation drift. It consists of Polestar-Cache for sparse cache refreshes and Polestar-Commit for identifying commit-ready tokens, achieving up to 10.73% accuracy improvement and 3.7x higher throughput on math and coding benchmarks.

This paper introduces token time continuous diffusion (TTCD), a diffusion language model operating in continuous space with per-token times, where tokens proceed from noise to token at varying rates. TTCD avoids parallel sampling inaccuracies and outperforms discrete models at high speedups. A 160M parameter model trained on OpenWebText and self-distilled achieves comparable unconditional and superior conditional generation, with gains in Sudoku solving.

The paper introduces AGOPS, an automatic method to generate task-specific prompt guidelines that help users write better prompts, improving LLM performance by recovering large performance drops from underspecification.

A paper presenting an educational system architecture that integrates UzWordnet and generative AI to enable Uzbek language practice through gaming, with four designed games and a methodology to enrich UzWordnet as a by-product.

A new study quantifies information loss when LLM agents communicate via text, using sparse autoencoder feature analysis. While latent communication preserves more information under compression, the lost features primarily encode surface form rather than task-relevant semantics, questioning the practical advantage of latent channels.

UniSAGE is a unified framework for modeling data with both static and dynamic attributes. It constructs a global attribute graph, introduces orthogonal parameter subspaces, and uses a hyper-structure mechanism for task-specific interactions. Experiments show over 10% improvement on multiple benchmarks.

This paper proposes LBA, a sampling-based method for generating high-quality adversarial texts under low query budgets in the hard-label setting. By integrating prior and posterior knowledge to construct an approximate distribution, LBA efficiently samples adversarial examples. Extensive experiments show LBA outperforms state-of-the-art baselines across models and datasets, with better semantic preservation and readability.

This paper presents the first application of pregroup grammar-based quantum compositional NLP to Arabic, a morphologically rich language. Quantum circuits mirror grammatical structure, outperforming classical baselines in word order, tense, and verb sense disambiguation experiments.

The Just Keep Prompting (JKP) framework tests VLM stability under repeated challenging. Evaluations on GPT-4o, Gemini 2.5 Pro, and Qwen3-VL-30B show substantial instability and answer flipping, with model-specific pressure-response profiles.

This paper analyzes why closed-loop knowledge systems (e.g., LLMs, RL) saturate under repeated internal feedback and introduces a three-level operational framework to enable escape via structural interventions. Using Lyapunov drift, stability is characterized, and escape is quantified by attractor displacement and a KL lower bound. Case studies include LLM code repair, sparse-reward RL, and Bayesian optimization.

Offline reinforcement learning world models suffer from model exploitation in low-data regions. RENEW uses human preferences over imagined rollouts to directly repair exploitation, introducing epistemic uncertainty to focus finetuning and improve sample efficiency.

This paper introduces an edge-aware learning-to-optimize framework for real-time relay selection in NR-V2X vehicular communications. By modeling V2X snapshots as directed graphs and using offline MILP solutions to supervise a Graph Isomorphism Network with Edge Features (GINE), the approach achieves inference latency within 5 ms. The GINE achieves 0.9589 accuracy and 0.9544 F1-score at link level, while a hybrid GP-MILP strategy reduces solver runtime below 30 ms for over 98% of graph instances while preserving optimality.

This study extends supervised lexicon learning to 10-K filings and their Item 1A risk-factor sections, training sentiment scores against return and volatility labels at three aggregation levels: sector, portfolio, and individual firm. Using 1,383 filings from 94 Nasdaq-100 technology firms (2006–2023), it finds full-text sentiment is more accurate at sector and portfolio levels, but Item 1A performs better at the individual firm level. A Loughran-McDonald dictionary baseline is consistently strongly negatively correlated with price, highlighting the value of a supervised approach for regulatory disclosure text.

Proposes Branching Policy Optimization (BPO), which leverages deterministic, snapshottable, and resumable sandboxes to construct a tree-structured rollout topology with shared prefixes, reducing policy gradient variance and improving success rates by 3.6–6.1 absolute points over GRPO and RLOO.

This paper introduces QFireNet, a quantum-hybrid U-Net model for wildfire segmentation from satellite imagery. Quantum-enhanced variants outperformed classical baselines on the Sen2Fire dataset, and data mixing significantly boosted performance by reducing domain shift.

This paper introduces C3R, a drop-in control layer that, from an inferred domain posterior and no query-time label, certifies a per-domain contamination budget where feasible and otherwise abstains. It guarantees a reduction on the hardest domains, shows stability across resampling, and retains more recall than calibrated cascades.