17 papers
Extreme Compression of Per-Layer Embeddings via Cross-Layer Latent Vector Quantization
Jiacheng Liu, Yaxin Luo, Xiaohan Zhao +5
Scaling embedding capacity alongside Transformer computation is an emerging trend in language models. This trend makes high compression rates increasingly useful. Hence, we study e…
DARTree: Speculative Diffusion Decoding with Autoregressive Draft Trees
Tianyi Li, Yaxin Luo, Xinyi Shang +1
Speculative decoding losslessly accelerates autoregressive language models by verifying multiple draft tokens in parallel. Diffusion-based drafters further reduce proposal latency…
Simple Domain Generalization for Strong Pixel-Level Image Tampering Detection in Modern VLMs
Yi Tang, Xinyi Shang, Jiacheng Cui +12
Modern vision-language models (VLMs) have significantly improved image generation and editing capabilities, making pixel-level image tampering detection increasingly important yet…
Three-Body Scattering for Generative Modeling
Peng Sun, Zhenglin Cheng, Deyuan Liu +3
Modern generative models typically rely on an adversarial critic, a prescribed noise-to-data path, or an autoregressive factorization. Instead, we show that a proper distributional…
Condensing Large-Scale Datasets Directly with Minimal Information Loss
Xinyi Shang, Peng Sun, Bei Shi +2
Recent advancements in scaling dataset distillation rely heavily on decoupled information extraction pipelines, comprising SQUEEZE, RECOVER, and RELABEL stages. Despite their scala…
Operation-Guided Progressive Human-to-AI Text Transformation Benchmark for Multi-Granularity AI-Text Detection
Sondos Mahmoud Bsharat, Jiacheng Liu, Xiaohan Zhao +9
As AI writing assistants become increasingly integrated into real-world drafting and revision workflows, many documents are no longer purely human-written or AI-generated, but inst…
LLMSurgeon: Diagnosing Data Mixture of Large Language Models
Yaxin Luo, Jiacheng Cui, Xiaohan Zhao +5
The pretraining data mixture of Large Language Models (LLMs) constitutes their "digital DNA", shaping model behaviors, capabilities, and failure modes. Yet this composition is rare…
Decision Boundary-aware Generation for Long-tailed Learning
Jiacheng Yang, Ruichi Zhang, Chikai Shang +5
Long-tailed data bias decision boundaries toward head classes and degrade tail class accuracy. Diffusion-based generative augmentation address this problem by generating additional…
Dive into Claude Code: The Design Space of Today's and Future AI Agent Systems
Jiacheng Liu, Xiaohan Zhao, Xinyi Shang +1
Claude Code is an agentic coding tool that can run shell commands, edit files, and call external services on behalf of the user. This study describes its architecture by analyzing…
From Masks to Pixels and Meaning: A New Taxonomy, Benchmark, and Metrics for VLM Image Tampering
Xinyi Shang, Yi Tang, Jiacheng Cui +9
Existing tampering detection benchmarks largely rely on object masks, which severely misalign with the true edit signal: many pixels inside a mask are untouched or only trivially m…
Exploring 3D Dataset Pruning
Xiaohan Zhao, Xinyi Shang, Jiacheng Liu +1
Dataset pruning has been widely studied for 2D images to remove redundancy and accelerate training, while particular pruning methods for 3D data remain largely unexplored. In this…
Fast and Scalable Analytical Diffusion
Xinyi Shang, Peng Sun, Jingyu Lin +1
Analytical diffusion models offer a mathematically transparent path to generative modeling by formulating the denoising score as an empirical-Bayes posterior mean. However, this in…
Duality Models: An Embarrassingly Simple One-step Generation Paradigm
Peng Sun, Xinyi Shang, Tao Lin +1
Consistency-based generative models like Shortcut and MeanFlow achieve impressive results via a target-aware design for solving the Probability Flow ODE (PF-ODE). Typically, such m…
Next-Gen CAPTCHAs: Leveraging the Cognitive Gap for Scalable and Diverse GUI-Agent Defense
Jiacheng Liu, Yaxin Luo, Jiacheng Cui +3
The rapid evolution of GUI-enabled agents has rendered traditional CAPTCHAs obsolete. While previous benchmarks like OpenCaptchaWorld established a baseline for evaluating multimod…
Equally Critical: Samples, Targets, and Their Mappings in Datasets
Runkang Yang, Peng Sun, Xinyi Shang +2
Data inherently possesses dual attributes: samples and targets. For targets, knowledge distillation has been widely employed to accelerate model convergence, primarily relying on t…
Collaborative Unlabeled Data Optimization
Xinyi Shang, Peng Sun, Fengyuan Liu +1
This paper pioneers a novel data-centric paradigm to maximize the utility of unlabeled data, tackling a critical question: How can we enhance the efficiency and sustainability of d…
GIFT: Unlocking Full Potential of Labels in Distilled Dataset at Near-zero Cost
Xinyi Shang, Peng Sun, Tao Lin
Recent advancements in dataset distillation have demonstrated the significant benefits of employing soft labels generated by pre-trained teacher models. In this paper, we introduce…