i’ve heard of the dataset poisoning and degradation caused by llm-generated content present in the dataset myself, but i’m not sure whether it was a practical observation, or a mere experiment. And I still fail to see how new datasets are really useful for developing a new llms, or how it’s a problem for the devs to switch back to the older datasets.
And the cornerstone stays the same: to have any significant effect on the final LLM quality, shouldn’t the poisoned (either by llm-produced content, or by intentional poisoning) data portion be… well, statistically significant?
i’ve heard of the dataset poisoning and degradation caused by llm-generated content present in the dataset myself, but i’m not sure whether it was a practical observation, or a mere experiment. And I still fail to see how new datasets are really useful for developing a new llms, or how it’s a problem for the devs to switch back to the older datasets.
And the cornerstone stays the same: to have any significant effect on the final LLM quality, shouldn’t the poisoned (either by llm-produced content, or by intentional poisoning) data portion be… well, statistically significant?