Reasoning Switch
An annotation dimension in post-training datasets used to distinguish whether a sample contains a chain of thought or internal reasoning process. Samples with reasoning retain the reasoning steps before the final answer, while samples without reasoning retain only a concise final response.
Timeline (last 90 days)
The reasoning switch (reasoning on/off) is an annotation dimension introduced by many post-training datasets: samples with reasoning retain the chain of thought before the final answer, while samples without reasoning retain only a concise final answer.
For customer service, shopping guide, and notification-type reply models, what is needed is the behavioral pattern of directly giving the answer; once a large number of reasoning-on samples are mixed into the training data, the model will treat first reasoning at length and then answering as the task specification.
All Facts (2)
The reasoning switch (reasoning on/off) is an annotation dimension introduced by many post-training datasets: samples with reasoning retain the chain of thought before the final answer, while samples without reasoning retain only a concise final answer.
50%UnverifiedFor customer service, shopping guide, and notification-type reply models, what is needed is the behavioral pattern of directly giving the answer; once a large number of reasoning-on samples are mixed into the training data, the model will treat first reasoning at length and then answering as the task specification.
50%