A Mamba-based per-token gate plus a Mamba-based paragraph selector improves frozen 7-9B LLMs' multi-hop QA by up to +9.3 F1 over LoRA, average +6.8 F1.
SSML o RA : Enhancing Low-Rank Adaptation with State Space Model
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Selective State-Space Adaptation and Retrieval for Language Model Reasoning
A Mamba-based per-token gate plus a Mamba-based paragraph selector improves frozen 7-9B LLMs' multi-hop QA by up to +9.3 F1 over LoRA, average +6.8 F1.