From Layers to Submodules: Rethinking Granularity in Replacement-Based LLM Compression
- Published
- Source
- arXiv
- Paper number
- 298
- Field
- LLMs / NLP
- arXiv ID
- 2606.02559
Key points
- Existing replacement-based methods share two design constraints: layer-level granularity and continuous selection.
- Building on this intuition, the authors introduce SubFit, or submodule-level fitted residual replacement, which compresses LLMs at the submodule level. Attention and feed-forward submodules are selected non-contiguously, and each receives its own lightweight fitted residual bypass path.
- SubFit operates after training and requires only calibration data.
Paper links
External research summaries. These are not HDATF publications or measured product results.