2 ms·
Keep in mind that tokens are LLM units of thought; the only moment the model does any computation is when generating tokens. Therefore, asking it to be succinct
by TeMPOraL 2y ago
Keep in mind that tokens are LLM units of thought; the only moment the model does any computation is when generating tokens. Therefore, asking it to be succinct means effectively dumbing it down.