Search papers, labs, and topics across Lattice.
Singapore University of Technology and Design
3
0
7
1
Removing misleading reasoning from answer-correct CoT traces can drastically enhance fine-tuning performance in LLMs.
The fragmented field of world modeling can now be unified under a "levels x laws" taxonomy, revealing critical gaps in autonomous model revision and decision-centric evaluation.
Scale up offline policy training for diffusion LLMs without breaking the bank: dTRPO slashes trajectory computation costs while boosting performance up to 9.6% on STEM tasks.
