Recency Is the Wrong Axis: A Negative Result on Windowed Deepest-Plane KV Caching in Looped Language Models
Tests a cache policy that retains every loop’s keys and values for recent positions and shares the deepest loop’s cache elsewhere. Across three looped language models, recent-position windows fail to provide a reliable repair; convergence across loops and which positions share matter more.
First public: · Added to this list: November 2, 2026
Earlier title: Prefix Keys and Values Matter Late: RIPE Caching for Self-Speculative Decoding in Looped Language Models.
Explore this approach
Recurrent reasoning — Repeat parts of the model or keep intermediate information in memory while working through a problem. Some methods also learn how many passes to run.
Analysis and evaluation — Measure how well these methods work, investigate what their internal computation does, and identify where they fail. This includes experiments, studies of model internals, and mathematical results.