Papers/2608.25068
🧪 Test?View on arXiv

SHIFT-LLM: Distribution Shift Correction in Depth-Pruned LLMs

Not specified in the provided content

depth pruningmodel compressionresidual networksfine-tuning
2608.25068
Builder Relevance
80%
2h ago

Abstract

SHIFT-LLM introduces a training-free post-pruning correction framework that mitigates accuracy loss in depth-pruned large language models.

Reality Card

Core Claim

SHIFT-LLM recovers accuracy lost to depth pruning by using Linear Residual Adapters, achieving gains up to +15.7 points on Llama-3.1-8B-Instruct.

Method / Result

Achieves accuracy recovery with only a few hundred calibration samples and no gradient computation.

Limitations

The method relies on a small held-out set for calibration, which may limit generalizability across different datasets.

Paper to code

Verified implementation resources so builders can test the paper’s claims instead of stopping at the abstract.

No verified implementation link has been attached yet. AIBuzzHub will keep this panel separate from unverified search results.
← Back to all papers