ECNU-SEA/

SEA-A

Paper Link: https://arxiv.org/abs/2407.12857

Project Page: https://ecnu-sea.github.io/

⚠️ This is the SEA-A model for assessing the consistency between paper content and reviews.

The SEA-A model leverages the pre-trained sentence representation model SFR-Embedding-Mistral to encode the paper text and the corresponding review text into vector representations. Based on these representations, the query and key vectors are computed independently for the paper and the review. These vectors are then fed into the SEA-A model to produce a mismatch score, which quantifies the degree of inconsistency between the paper content and the review.