AI Paper Atlas

Research map / Architectures and efficiency / Pruning and efficient tuning

HiFICL: High-Fidelity In-Context Learning for Multimodal Tasks

Xiaoyu Li, Yuhang Liu, Xuanshuo Kang, Zheng Luo, Fangqi Lou, Xiaohua Wu, Zihan Xiong

CVPR 2026 · Architectures and efficiency › Pruning and efficient tuning · approach: Parameter-efficient transfer analysis · 0 citations (OpenAlex)

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
HiFICL virtual key-value pairs instead of MimIC shift vectorLLaVA-Interleave-7B / Idefics2 · OK-VQA, VQAv2VQA accuracy: OK-VQA 52.29 -> 54.19 (LLaVA); VQAv2 69.29 -> 72.08 (Idefics2)Improved
HiFICL (2.2M params) versus LoRA/MimIC on LLaVALLaVA-Interleave-7B · VQAv2trainable params and accuracy: 17.0M -> 2.2M params; 74.40 -> 74.66Improved
increase ICL shots from 16 to 32Idefics2 ICL · VQAv2VQA accuracy: 67.70 -> 50.55Hurt
add teacher-student alignment to HiFICL (ablation)HiFICL · VQAv2VQA accuracy: 72.08 -> 70.09Hurt
add low-rank factorization on key matrix (ablation)HiFICL · OK-VQAVQA accuracy: 55.72 -> 59.56Improved
add low-rank factorization on value matrix (ablation)HiFICL · VQAv2VQA accuracy: 69.31 -> 72.08Improved
add self-attention scaling alpha (non-linear modulation) (ablation)HiFICL · VQAv2VQA accuracy: 70.14 -> 72.08Improved
teacher-free training instead of teacher (MimIC) paradigm: training costHiFICL · VQAv2 training costrelative FLOPs: 14.3x (MimIC) -> 1xImproved
rank r=16 instead of 8 (ablation)HiFICL · OK-VQA vs VQAv2 (Idefics2)accuracyMixed
HiFICL instead of 8-shot ICL on COCO captionsIdefics2 · COCO hallucinationCHAIRi: 3.9 -> 2.2Improved

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