Search papers, labs, and topics across Lattice.
1
0
3
Skip the expensive proxy model training: this training-free method boosts VLLM performance by up to 4.8% using only 10-15% of the data, simply by measuring how much the question *changes* the model's view of the answer.
