A survey synthesizing LLM methods for peer review critique generation and score prediction, including taxonomies, benchmark limitations, domain biases, and robustness risks such as prompt injection.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
LLM-Based Scientific Peer Review: Methods, Benchmarks, and Reliability Challenges
A survey synthesizing LLM methods for peer review critique generation and score prediction, including taxonomies, benchmark limitations, domain biases, and robustness risks such as prompt injection.