Search papers, labs, and topics across Lattice.
Affiliation:
3
0
5
2
Humor comprehension in AI just got a boost, with CaRGo-T achieving up to 20% better performance by leveraging causal reasoning graphs.
ECI_{sem} ranks LLM negatives highest among non-hybrid sources, revealing that effective hard-negative selection can be achieved without fine-tuning.
You can accurately predict the NDCG of a 1B-parameter reranking model by only training models up to 400M parameters, unlocking massive compute savings.
