Poisoning's two strongest results were never run in the same experiment
Sleeper Agents held a hidden trigger through the full safety stack, and a separate study planted one with just 250 documents, yet neither experiment did what the other did.
A perfect needle-in-a-haystack score means the model found one planted sentence
Gemini 1.5 clears the test through a million tokens, a benchmark that rewards copying an out-of-place fact and leaves long-context reasoning untested.
A model ends its reply by choosing a stop token like any other word
The model learns that stop token in post-training, and a separate length cap can cut a reply off mid-word before it ever arrives.
A health algorithm predicted cost, and scored sicker Black patients as lower-risk
A 2019 study of nearly 50,000 patients showed the tool doing exactly what it was built to do, predicting next year's medical spending rather than illness.
king - man + woman = queen scored 65.6% on word2vec's own test
Tomas Mikolov and colleagues at Google reported the figure on a 19,544-question benchmark in 2013, and later analyses found the vector offset leans on excluding the very words the analogy starts from.