๐ŸŽ“How I Study AIHISA
๐Ÿ“–Read
๐Ÿ“„Papers๐Ÿ“ฐBlogs๐ŸŽฌCourses
๐Ÿ’กLearn
๐Ÿ›ค๏ธPaths๐Ÿ“šTopics๐Ÿ’กConcepts๐ŸŽดShorts
๐ŸŽฏPractice
๐ŸงฉProblems๐ŸŽฏPrompts๐Ÿง Review
Search
How I Study AI - Learn AI Papers & Lectures the Easy Way

Papers2

AllBeginnerIntermediateAdvanced
All SourcesarXiv
#approximate nearest neighbor

Fast Autoregressive Video Diffusion and World Models with Temporal Cache Compression and Sparse Attention

Intermediate
Dvir Samuel, Issar Tzachor et al.Feb 2arXiv

The paper makes long video generation much faster and lighter on memory by cutting out repeated work in attention.

#autoregressive video diffusion#KV cache compression#sparse attention

Reveal Hidden Pitfalls and Navigate Next Generation of Vector Similarity Search from Task-Centric Views

Beginner
Tingyang Chen, Cong Fu et al.Dec 15arXiv

The paper shows that judging vector search only by distance-based recall and speed can be very misleading for real tasks.

#vector similarity search#approximate nearest neighbor#maximum inner product search