Skip to content
Live
Trending storyPast story

Apple proposed a new method to improve the efficiency of long-context retrieval.

1 reports1 outlets9 days agoUpdate

The story so far

AI overview

Apple Machine Learning Research proposes a Weighted State Souping method that combines cached State state encodings using query dependency weights, significantly improving long-context performance. This method is suitable for tasks such as single-hop and multi-hop document question answering, pinpoint retrieval, variable tracking, and tool use, and can be adapted into an efficient sparse variant to achieve near-constant encoding costs.

AI-written from the reports · Updated 7 days ago

Coverage timeline

Follow the coverage to see every side of the story.

10 Nov
  1. Apple Machine Learning Research
    Weighted State Souping for Efficient Long-Context Retrieval

    Weighted State Souping proposes query dependency weighting to combine cached State state encodings, significantly improving long-context performance on tasks such as single-hop and multi-hop document question answering, pinpoint retrieval, variable tracking, and tool use. The method can be adapted into an efficient sparse variant, achieving near-constant encoding costs while maintaining (or sometimes improving) performance.

Heat over time

Not enough data yet to draw a trend.