Context compression finally works in production: new research cuts LLM input 16x without the accuracy hit

LCLMs compress LLM context before decode — 8.8x faster at 16x compression, beating every KV cache method tested. Open-sourced by NYU and Columbia.
IEEE

JND-LIC: Learned Image Compression via Just Noticeable Difference for Human Visual Perception

Abstract: Existing human visual perception-oriented image compression methods well maintain the perceptual quality of compressed images, but they may introduce fake details into the compressed images, ...
IEEE

Learned Image Compression via Local-to-Global Cross-Component Prior

Abstract: Learned image compression (LIC) methods have shown promising results and achieved superior performance compared to traditional image compression methods. Due to the neglect of the ...
GitHub

TokenSkip: Controllable Chain-of-Thought Compression in LLMs

Does every token in the CoT output contribute equally to deriving the answer? —— We say NO! We introduce TokenSkip, a simple yet effective approach that enables LLMs to selectively skip redundant ...
GitHub

cctx — Claude Code Token Optimizer

Claude Code's context window fills up fast. Every bash run, every file read, every grep result gets appended verbatim. By turn 15 you're paying for — and waiting on — tens of thousands of tokens of ...