1 citations · 2 across the 11 of their papers we have counts for
14 papers
Residual-Stream Burden Shapes Representation Learning in Diffusion Transformers
Tongtong Liang, Siqi Kou, Ziqiao Xi +6
In diffusion-based generation, a neural network can be trained to predict the clean data, the noise, or the velocity from a noisy input. These prediction targets are interconvertib…
MirrorPPR: Exemplar-Based Portrait Photo Retouching
Zhihong Liu, Zheng Li, Jiachun Jin +4
While text-guided image editing has made remarkable progress, it remains limited in structural portrait retouching. Textual descriptions struggle to convey fine-grained changes to…
World-Language-Action Model for Unified World Modeling, Language Reasoning, and Action Synthesis
Yi Yang, Zhihong Liu, Siqi Kou +9
We propose world-language-action (WLA) models as a new class of embodied foundation models. WLA takes textual instructions, images, and robot states as inputs to jointly predict te…
ProductWebGen: Benchmarking Multimodal Product Webpage Generation
Zhihong Liu, Siqi Kou, Zheng Li +5
Crafting a product display webpage from a source product image, along with layout and visual content instructions, holds significant practical value for domains such as marketing,…
Think-Then-Generate: Reasoning-Aware Text-to-Image Diffusion with LLM Encoders
Siqi Kou, Jiachun Jin, Zetong Zhou +8
Recent progress in text-to-image (T2I) diffusion models (DMs) has enabled high-quality visual synthesis from diverse textual prompts. Yet, most existing T2I DMs, even those equippe…
Fast and Accurate Causal Parallel Decoding using Jacobi Forcing
Lanxiang Hu, Siqi Kou, Yichao Fu +5
Multi-token generation has emerged as a promising paradigm for accelerating transformer-based large model inference. Recent efforts primarily explore diffusion Large Language Model…
LoHoVLA: A Unified Vision-Language-Action Model for Long-Horizon Embodied Tasks
Yi Yang, Jiaxuan Sun, Siqi Kou +2
Real-world embodied agents face long-horizon tasks, characterized by high-level goals demanding multi-step solutions beyond single actions. Successfully navigating these requires b…
Thinking with Generated Images
Ethan Chern, Zhulin Hu, Steffi Chern +5
We present Thinking with Generated Images, a novel paradigm that fundamentally transforms how large multimodal models (LMMs) engage with visual reasoning by enabling them to native…
Which Data Attributes Stimulate Math and Code Reasoning? An Investigation via Influence Functions
Siqi Kou, Qingyuan Tian, Hanwen Xu +2
Large language models (LLMs) have demonstrated remarkable reasoning capabilities in math and coding, often bolstered by post-training on the chain-of-thoughts (CoTs) generated by s…
Orthus: Autoregressive Interleaved Image-Text Generation with Modality-Specific Heads
Siqi Kou, Jiachun Jin, Zhihong Liu +6
We introduce Orthus, an autoregressive (AR) transformer that excels in generating images given textual prompts, answering questions based on visual inputs, and even crafting length…
MatryoshkaKV: Adaptive KV Compression via Trainable Orthogonal Projection
Bokai Lin, Zihao Zeng, Zipeng Xiao +5
KV cache has become a de facto technique for the inference of large language models (LLMs), where tensors of shape (layer number, head number, sequence length, feature dimension) a…
CLLMs: Consistency Large Language Models
Siqi Kou, Lanxiang Hu, Zhezhi He +2
Parallel decoding methods such as Jacobi decoding show promise for more efficient LLM inference as it breaks the sequential nature of the LLM decoding process and transforms it int…
BayesDiff: Estimating Pixel-wise Uncertainty in Diffusion via Bayesian Inference
Siqi Kou, Lei Gan, Dequan Wang +2
Diffusion models have impressive image generation capability, but low-quality generations still exist, and their identification remains challenging due to the lack of a proper samp…
Phasic Content Fusing Diffusion Model with Directional Distribution Consistency for Few-Shot Model Adaption
Teng Hu, Jiangning Zhang, Liang Liu +7
Training a generative model with limited number of samples is a challenging task. Current methods primarily rely on few-shot model adaption to train the network. However, in scenar…