3 ms·
Using predictions of past events as the benchmark where the LLMs have already been trained on the results seems quite flawed to me
by unknownx113 7mo ago
Using predictions of past events as the benchmark where the LLMs have already been trained on the results seems quite flawed to me