5318 papers
This paper introduces a unified language modeling approach for semantic and panoptic segmentation in images and videos that discretizes masks into run-length encoded tokens, employing novel compression strategies to enable autoregressive generation despite computational constraints.
This paper introduces EmoOmni, a unified framework that leverages an emotional Chain-of-Thought (E-CoT) to bridge the gap between fine-grained multimodal perception and accurate emotional expression in Omni-LLMs, accompanied by a new dataset and benchmark for systematic evaluation.
CryoNet.Refine is a novel one-step diffusion model that automates and accelerates the refinement of atomic structures against cryo-EM density maps, outperforming traditional tools like Phenix in both model-map correlation and geometric quality while supporting diverse protein and nucleic acid complexes.
This paper argues that AI agents equipped with specialized skills can augment, but not fully replace, social scientists by executing codifiable research tasks autonomously through "vibe researching," while highlighting the enduring necessity of human theoretical originality and tacit knowledge alongside the profession's emerging risks of stratification and pedagogical crisis.
This paper analyzes physician disagreement in the HealthBench dataset, revealing that while the majority of variance is structural and irreducible, a small but actionable portion stems from reducible uncertainties like missing context, suggesting that improving evaluation design to close information gaps could meaningfully reduce disagreement on borderline medical AI cases.
WISER is a training-free framework for Zero-Shot Composed Image Retrieval that unifies Text-to-Image and Image-to-Image paradigms through a "retrieve-verify-refine" pipeline, leveraging wider search, adaptive fusion, and self-reflection to significantly outperform existing methods across diverse benchmarks.
The paper introduces PackUV, a novel 4D Gaussian representation and fitting method that maps volumetric video attributes into structured UV atlases for efficient, codec-compatible storage and streaming, while demonstrating superior temporal consistency and rendering fidelity on the newly proposed large-scale PackUV-2B dataset.
This paper proposes a sample-efficient approach to generalized planning that learns explicit neural transition models to predict intermediate world states, demonstrating superior out-of-distribution performance and data efficiency compared to direct action-sequence prediction methods.
This paper proposes HART, an annotation-free framework that leverages a novel Advantage Preference Group Relative Policy Optimization (AP-GRPO) algorithm to enable Large Multimodal Models to autonomously identify and verify key high-resolution image regions, thereby improving reasoning performance without requiring costly human grounding labels.
This paper introduces PEPA, a three-layer cognitive architecture that leverages personality traits to enable embodied agents to autonomously generate goals and sustain long-term operation in dynamic environments without relying on external task specifications.
This paper introduces Infinite Self-Attention (InfSA) and its linear-time variant, Linear-InfSA, a spectral reformulation of self-attention as a diffusion process on token graphs that achieves state-of-the-art ImageNet accuracy and enables efficient, memory-free inference at ultra-high resolutions (up to 9216×9216) by replacing the quadratic softmax cost with a Neumann series approximation.
This paper introduces WildActor, a framework for unconstrained identity-preserving human video generation that leverages the large-scale Actor-18M dataset and novel attention mechanisms to overcome existing limitations in maintaining consistent full-body identities across dynamic shots, viewpoints, and motions.
This position paper argues that the evaluation of modern visual processing systems must shift from a reliance on single-metric benchmarks toward a human-centered, context-aware paradigm to better align with human perception and foster genuine innovation.
This workshop paper proposes "sustainable care" as a design framework to help researchers and practitioners create digital technologies that foster children's long-term, meaningful engagement with social issues while preventing empathic distress and burnout.
The paper proposes DeAR, a fine-grained adaptation framework for Vision-Language Models that decomposes attention heads into functional roles (Attribute, Generalization, and Mixed) using a Concept Entropy metric to selectively isolate task-specific learning from generalization capabilities, thereby achieving superior performance across diverse tasks while preserving zero-shot robustness.
This paper presents a validated digital twin of the Frontier supercomputer's liquid cooling system to demonstrate that a ramp-constrained, joint optimization of flow rate and supply temperature can achieve 27.8% energy savings, significantly outperforming flow-only strategies by addressing the gap between theoretical optima and operational deployability.
This technical report extends the empirical validation of the Explainability Solution Space (ESS) framework by demonstrating its domain-independent applicability and systematic adaptability to diverse governance roles and stakeholder configurations through a cross-domain evaluation involving both employee attrition and urban resource allocation systems.
This paper proposes and evaluates static and adaptive traffic scheduling schemes—including threshold, heuristic, and reinforcement learning-based approaches—to optimize energy efficiency and delivery ratios for energy-constrained optical LEO satellite downlinks facing weather-related disruptions.
HarmonyCell is an end-to-end agent framework that automates single-cell perturbation modeling by combining an LLM-driven semantic unifier to resolve metadata incompatibilities and an adaptive Monte Carlo Tree Search engine to synthesize architectures that handle distribution shifts, thereby achieving high execution success and outperforming expert baselines without manual engineering.
This paper proposes a novel LLM-driven closed-loop framework that maps natural language instructions to executable rules and semantically annotates options to enhance the data efficiency, interpretability, and cross-environment transferability of Deep Reinforcement Learning, with experimental validation showing superior performance in constraint compliance and skill reuse.