A learned continue-thinking token, trained via RL on its embedding alone, improves math benchmark accuracy more than fixed-token budget forcing in a frozen language model.
Figure 6: GSM8K reasoning trace demonstrating the positive impact of the <|continue_thinking|> token
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2025 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Learning a Continue-Thinking Token for Enhanced Test-Time Scaling
A learned continue-thinking token, trained via RL on its embedding alone, improves math benchmark accuracy more than fixed-token budget forcing in a frozen language model.