3 ms·
Cool trick. But is this better than reinforcement learning, where the LLM decides for itself the optimal thinking time for each prompt?
by vagab0nd 2y ago
Cool trick. But is this better than reinforcement learning, where the LLM decides for itself the optimal thinking time for each prompt?