attention analysis 1entity alignment 1knowledge graphs 1large language models 1patch-level semantics 1representation learning 1training-free methods 1vision-language models 1zero-shot matching 1zero-shot segmentation 1
From the 2 of 2 linked papers with an AI index.
2 papers
cs.CV2026
TraceCLIP: Recovering Local Semantics from Patch-to-CLS Contributions
Xinran Liu, Shouqian Shi, Yutong Chen +3
The paper presents TraceCLIP, a training‑free method that extracts patch‑level semantic information from CLIP's CLS attention output to improve zero‑shot dense vision‑language task…
cs.CL2026
IRIS: Reusable Identity Representations from Frozen LLMs for Entity Alignment
Xinran Liu, Shengtao Li, Shouqian Shi +2
The paper introduces IRIS, a training-free method that uses frozen large language models to generate stable identity signatures for entities, enabling direct similarity-based align…