2 citations · 3 across the 15 of their papers we have counts for
17 papers
CoVisco: Codec-Native Vision Encoder with Native Token Compression for Unified Image-Video Understanding
Yulong Liu, Xiaotian Han, Junyuan Shang +6
Vision-language models face a fundamental scaling bottleneck: the number of visual tokens grows with both temporal duration and spatial resolution, making long-video understanding…
RESUME: Recurrent State Updates from Motion and Residual Signals for Efficient Video Language Modeling
Can Zhang, Xiaotian Han, Junyuan Shang +5
Existing video language models encode sampled RGB frames independently, so a long video must either exhaust the token budget or drop the changes between sampled frames. Codec-aware…
From Intent to Evidence: Policy-Steered Multi-Strategy Retrieval for Long-Video Agents
Can Zhang, Baofeng Zhang, Xiaotian Han +5
Existing long-video agents acquire evidence through one uniform behavior, ignoring whether the required evidence is concentrated, requires broad occurrence coverage, or must discri…
Memory Augmentation Unlocks Efficient Chain-of-Thought Reasoning
Simeng Zhang, Yilong Chen, Wenyuan Zhang +4
Large language models often rely on Chain-of-Thought (CoT) reasoning to solve complex tasks, but verbose reasoning traces introduce substantial inference overhead. CoT compression…
Aggregating Visual Information with Optimal Transport for VideoLM Token Compression
Wenti Yin, Xiaotian Han, Junyuan Shang +5
Video language models process videos as dense visual-token sequences with substantial representational redundancy. Compressing these sequences is therefore essential for reducing t…
Autonomy-of-Heads: Data-Free Sparse Attention from Frozen Query-Key Geometry
Yehan Yang, Junyuan Shang, Yang Li +3
Long-context LLM inference is bottlenecked by quadratic attention computation and growing KV-cache costs. Existing sparse attention and KV-compression methods typically decide whic…
Revisiting Lossy Verification in Speculative Decoding: Mechanisms, Trade-offs, and Failure Modes
Tianyu Wang, Yuxuan Zhou, Heng Li +4
Speculative Decoding (SD) accelerates large language model inference by allowing a lightweight draft model to propose tokens that are subsequently verified in parallel by a larger…
ConSA: Controllable Sparsity in Hybrid Attention via Learnable Allocation
Yao Chen, Yinqi Yang, Junyuan Shang +6
Hybrid architectures combining full attention (FA) and sliding-window attention (SWA) are a promising paradigm for efficient LLM inference. However, existing methods typically rely…
Sparse Growing Transformer: Training-Time Sparse Depth Allocation via Progressive Attention Looping
Yao Chen, Yilong Chen, Yinqi Yang +9
Existing approaches to increasing the effective depth of Transformers predominantly rely on parameter reuse, extending computation through recursive execution. Under this paradigm,…
Mixture of Universal Experts: Scaling Virtual Width via Depth-Width Transformation
Yilong Chen, Naibin Gu, Junyuan Shang +8
Mixture-of-Experts (MoE) decouples model capacity from per-token computation, yet their scalability remains limited by the physical dimensions of depth and width. To overcome this,…
ERNIE 5.0 Technical Report
Haifeng Wang, Hua Wu, Tian Wu +432
In this report, we introduce ERNIE 5.0, a natively autoregressive foundation model desinged for unified multimodal understanding and generation across text, image, video, and audio…
VideoAR: Autoregressive Video Generation via Next-Frame & Scale Prediction
Longbin Ji, Xiaoxiong Liu, Junyuan Shang +4
Recent advances in video generation have been dominated by diffusion and flow-matching models, which produce high-quality results but remain computationally intensive and difficult…
Inner Thinking Transformer: Leveraging Dynamic Depth Scaling to Foster Adaptive Internal Thinking
Yilong Chen, Junyuan Shang, Zhenyu Zhang +7
Large language models (LLMs) face inherent performance bottlenecks under parameter constraints, particularly in processing critical tokens that demand complex reasoning. Empirical…
Mixture of Hidden-Dimensions Transformer
Yilong Chen, Junyuan Shang, Zhengyu Zhang +6
Transformer models encounter challenges in scaling hidden dimensions efficiently, as uniformly increasing them inflates computational and memory costs while failing to emphasize th…
MoR: Mixture of Ranks for Low-Rank Adaptation Tuning
Chuanyu Tang, Yilong Chen, Zhenyu Zhang +4
Low-Rank Adaptation (LoRA) drives research to align its performance with full fine-tuning. However, significant challenges remain: (1) Simply increasing the rank size of LoRA does…
NACL: A General and Effective KV Cache Eviction Framework for LLMs at Inference Time
Yilong Chen, Guoxia Wang, Junyuan Shang +7
Large Language Models (LLMs) have ignited an innovative surge of AI applications, marking a new era of exciting possibilities equipped with extended context windows. However, hosti…
DHA: Learning Decoupled-Head Attention from Transformer Checkpoints via Adaptive Heads Fusion
Yilong Chen, Linhao Zhang, Junyuan Shang +4
Large language models (LLMs) with billions of parameters demonstrate impressive performance. However, the widely used Multi-Head Attention (MHA) in LLMs incurs substantial computat…