Search papers, labs, and topics across Lattice.
Affiliation:
3
0
4
0
Resource-aware LLM task graphs can achieve competitive accuracy with 3.17x less cost and latency by jointly executing and verifying subtasks, learning LLM quality online.
Naive importance weighting falls apart in deep hierarchical inference systems with sparse feedback, but a variance-reduced EXP4 algorithm can restore stability and near-optimal routing.
SPRINT tackles the overlooked challenge of few-shot class-incremental learning on tabular data, achieving state-of-the-art results by smartly exploiting unlabeled data and cheap storage.
