MIDUS replaces duplicated FFN branches in depth up-scaling with head-wise memory layers using product-key retrieval and HIVE to deliver lightweight, head-conditioned residual capacity.
Focus direc- tions make your language models pay more attention to relevant contexts.arXiv preprint arXiv:2503.23306
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
verdicts
UNVERDICTED 2representative citing papers
EASE-TTT creates a soft attention target from evidence chunks to guide query-side test-time adaptation, yielding higher macro-average scores than full-context, retrieval-only, and standard qTTT baselines on six LongBench QA tasks.
citing papers explorer
-
MIDUS: Memory-Infused Depth Up-Scaling
MIDUS replaces duplicated FFN branches in depth up-scaling with head-wise memory layers using product-key retrieval and HIVE to deliver lightweight, head-conditioned residual capacity.
-
EASE-TTT: Evidence-Aligned Selective Test-Time Training for Long-Context Question Answering
EASE-TTT creates a soft attention target from evidence chunks to guide query-side test-time adaptation, yielding higher macro-average scores than full-context, retrieval-only, and standard qTTT baselines on six LongBench QA tasks.