1 paper · 1 filter
Haowei Lin, Hubert Lim, Xiangyu Wang +2
Test-time scaling (TTS) improves language model reasoning by allocating additional test-time compute, but prior work mainly studies closed-ended tasks such as math and coding. We s…