arXiv Computation and Language By Stanley Z. Hua, Sanae Lotfi, Irene Y. Chen

Investigating Social Bias Changes in Quantized Language Models

Read the original on arXiv Computation and Language →

Post‑training quantization of large language models reduces memory usage but can alter social biases in ways that aggregate metrics miss. In a large‑scale study of 50 quantized models on PostTrainingBiasBench, the authors discovered a phenomenon called quantization‑induced bias flipping, where up to 21% of responses switch from biased to unbiased or vice versa, especially for uncertain predictions and stronger quantization (4‑bit vs 8‑bit). These flips lead to asymmetric impacts across demographic groups, with some groups experiencing up to an 18.6% worsening of bias while others improve by 14.1%, resulting in misleadingly neutral overall scores.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv Computation and Language.

arXiv Computation and Language
Aug 21

The Asymmetric Harms of LLM Compression

arXiv:2608. 19670v1 Announce Type: new Abstract: Large language models (LLMs) compression reduces deployment costs, but standard aggregate metrics like perplexity and accuracy often mask underlying behavioral shifts.

By Yuan Wu, Mairui Li, Lesia Semenova, Chudi Zhong