It's because BLOOM is undertrained, you can prune a lot of weights in BLOOM and it doesn't impact performance. Look at Chinchilla paper[1], 70B model outperforms 175B GPT-3 model.
In general, most giant LLMs are extremely undertrained at this time. Consider that most of the gains in RoBerta vs bert were from just continuing to train.
https://arxiv.org/abs/2203.15556