LLM Digest
Subscribe

Story

arxiv_cs_cl · Nov 18, 2026 · paper

Source brief

Induction Head Origins of SLAM-ASR Speech Recognition Models

arxiv.orgNov 18, 2026
original source linked

In brief

What makes a small trained map from encoder states to the embedding space of a frozen language model sufficient for competitive speech recognition? The premise of the SLAM-ASR architecture is that a frozen language model possesses linguistic knowledge rich enough to perform the r

Continue reading

Read the original at arxiv.org →Open in live feedRead that day’s brief

Earlier in this thread 4 items