Story

arxiv_cs_cl ยท Oct 30, 2026 ยท paper

Source brief

DecisionRank: On-Policy Distillation for Decision-Model Rerankers

arxiv.orgOct 30, 2026
original source linked

In brief

Decision models target structured tasks by directly predicting choices or scores, avoiding autoregressive decoding latency. However, transferring a generative LLM's ranking knowledge to a decision model remains challe...

Feed lens
eval

Continue reading

Read the original at arxiv.org โ†’Open in live feedRead that dayโ€™s brief

Earlier in this thread 4 items