Pruning

Decoding Compressed Trust: Scrutinizing the Trustworthiness of Efficient LLMs Under Compression

A comprehensive trustworthiness assessment of compressed LLMs.