SL
Shiming Liu
cs.CVcs.AIcs.CLcs.LG
On Valency
published · living versionsW_zjawfm7v·v1 · currentpublished
ECHO: Emergent Common-Mode Hidden Offsets Drive Vocabulary Collapse in Preference Optimization
with Yun Li, Ehsan Javanmardi, Simon Thompson, Jiachen Gong +5
1 version
Preprints & journals
5 papers in the corpus · 2025–2026ECHO: Emergent Common-Mode Hidden Offsets Drive Vocabulary Collapse in Preference Optimization2610.29410v1 · Yun Li, Ehsan Javanmardi, Simon Thompson et al.2026 · 0 citationsarXivon Valency
Consistent Evidence, Robust Recognition: Faithful Attribution Regularization under Geometric Transformations2607.23835v1 · Xianghao Jiao, Ruoyu Chen, Wei Wang et al.2026 · 0 citationsarXiv
Where Not to Learn: Prior-Aligned Training with Subset-based Attribution Constraints2602.07008v6 · Ruoyu Chen, Shangquan Sun, Xiaoqing Guo et al.2026 · 0 citationsarXiv
Evidence Recomposition and Predictive Context Residualization for Visual Attribution in Multimodal Large Language Models2509.22415v5 · Jiawei Liang, Jianjie Huang, Xianghao Jiao et al.2025 · 0 citationsarXiv
Where MLLMs Attend and What They Rely On: Explaining Autoregressive Token Generation2509.22496v5 · Ruoyu Chen, Xiaoqing Guo, Kangwei Liu et al.2025 · 0 citationsarXiv
Career total: 6 works. 5 are in this corpus.
Profile built from the corpus for this byline.
Author records are still filling in while the Hub is in alpha. If this is your page, you'll be able to claim it soon. Spot a mistake? Tell us.