会员服务
Efficient LLM Inference using Dynamic Input Pruning and Cache-Aware Masking
Arxiv
1+阅读 · 4月3日
Efficient Autoregressive Shape Generation via Octree-Based Adaptive Tokenization
Efficient Autoregressive Shape Generation via Octree-Based Adaptive Tokenization
Arxiv
0+阅读 · 4月3日
Overcoming Vocabulary Constraints with Pixel-level Fallback
Overcoming Vocabulary Constraints with Pixel-level Fallback
Arxiv
0+阅读 · 4月2日
TokenHSI: Unified Synthesis of Physical Human-Scene Interactions through Task Tokenization
TokenHSI: Unified Synthesis of Physical Human-Scene Interactions through Task Tokenization
Arxiv
0+阅读 · 4月3日
Improving Applicability of Deep Learning based Token Classification models during Training
Arxiv
0+阅读 · 3月28日
PIM-LLM: A High-Throughput Hybrid PIM Architecture for 1-bit LLMs
Arxiv
0+阅读 · 3月31日
Why do LLMs attend to the first token?
Arxiv
0+阅读 · 4月3日
Ichigo: Mixed-Modal Early-Fusion Realtime Voice Assistant
Ichigo: Mixed-Modal Early-Fusion Realtime Voice Assistant
Arxiv
0+阅读 · 4月3日
Fine-Tuning Visual Autoregressive Models for Subject-Driven Generation
Arxiv
0+阅读 · 4月3日
Token Dynamics: Towards Efficient and Dynamic Video Token Representation for Video Large Language Models
Arxiv
0+阅读 · 4月2日
Mixtera: A Data Plane for Foundation Model Training
Mixtera: A Data Plane for Foundation Model Training
Arxiv
0+阅读 · 4月3日
The Foundations of Tokenization: Statistical and Computational Concerns
Arxiv
0+阅读 · 4月3日
ILLUME+: Illuminating Unified MLLM with Dual Visual Tokenization and Diffusion Refinement
Arxiv
0+阅读 · 4月3日
Parallelized Autoregressive Visual Generation
Arxiv
0+阅读 · 4月3日
Accelerate Parallelizable Reasoning via Parallel Decoding within One Sequence
Arxiv
0+阅读 · 4月2日
参考链接
微信扫码咨询专知VIP会员
Top