LLM-based dense retrievers generalize better when instruction-tuned but pay a specialization tax when optimized for reasoning; they resist typos and corpus poisoning better than encoder-only baselines yet remain vulnerable to semantic perturbations, with larger models and certain embedding geometry,
Shane Culpepper, Guglielmo Faggioli, Nicola Ferro, and Oren Kurland
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
method 1
citation-polarity summary
fields
cs.IR 1years
2026 1verdicts
UNVERDICTED 1roles
method 1polarities
use method 1representative citing papers
citing papers explorer
-
On the Robustness of LLM-Based Dense Retrievers: A Systematic Analysis of Generalizability and Stability
LLM-based dense retrievers generalize better when instruction-tuned but pay a specialization tax when optimized for reasoning; they resist typos and corpus poisoning better than encoder-only baselines yet remain vulnerable to semantic perturbations, with larger models and certain embedding geometry,