AI Paper Atlas

Research map / Vision-language and multimodal / Cross-modal learning

LIMSSR: LLM-Driven Sequence-to-Score Reasoning under Training-Time Incomplete Multimodal Observations

Huangbiao Xu, Huanqi Wu, Xiao Ke, Yuxin Peng

ICML 2026 · Vision-language and multimodal › Cross-modal learning · approach: Multimodal fusion with missing modalities

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add LLM backbone to baseline (ablation)LIMSSR baseline · FS1000 averageSpearman correlation: 0.528 -> 0.601Improved
add PCMI prompt-guided imputation (ablation)LIMSSR · FS1000 averageSpearman correlation: 0.601 -> 0.687Improved
add LMRF fusion (ablation)LIMSSR · FS1000 averageSpearman correlation: 0.687 -> 0.768Improved
add MDA dual-path aggregation (ablation)LIMSSR · FS1000 averageMSE: 18.18 -> 14.08Improved
add LMRF (ablation)LIMSSR · FS1000 averageSpearman correlation: 0.732 -> 0.789Improved
add PCMI (ablation)LIMSSR · Fis-V averageSpearman correlation: 0.699 -> 0.743Improved
add contrastive loss L_con (ablation)LIMSSR · FS1000 averageSpearman correlation: 0.759 -> 0.789Improved
add regularization loss L_reg (ablation)LIMSSR · FS1000 averageSpearman correlation: 0.773 -> 0.789Improved
add both MDA paths instead of Path 1 only (ablation)LIMSSR · FS1000 averageSpearman correlation: 0.778 -> 0.789Improved
add all prompt components (ablation)LIMSSR · FS1000 averageSpearman correlation: 0.710 -> 0.789Improved
increase fusion tokens K from 2 to 4 (ablation)LIMSSR · FS1000 averageSpearman correlation: 0.762 -> 0.789Improved
increase fusion tokens K from 4 to 6 (ablation)LIMSSR · FS1000 averageSpearman correlation: 0.789 -> 0.779Hurt

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy