AI Paper Atlas

Research map / Vision-language and multimodal / Multimodal LLMs

Improving Large Vision and Language Models by Learning from a Panel of Peers

Jefferson Hernandez, Jing Shi, Simon Jenni, Vicente Ordonez, Kushal Kafle

ICCV 2025 · Vision-language and multimodal › Multimodal LLMs · approach: VLM capability analysis · 1 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
run three iterations of self-improvement instead of onePoP-Vicuna · 15 vision-language benchmarks (average)average score: 53.1 -> 56.7Improved
run three iterations instead of onePoP LLaVA-1.5 · ScienceQAaccuracy: 71.2 -> 86.4Improved
use full panel instead of single-try panel (st-PoP)PoP LLaVA-1.5 iter3 · MMBenchaccuracy: 67.4 -> 72.5Improved
use full panel instead of single-try panel (st-PoP)PoP LLaVA-1.5 iter3 · POPEaccuracy: 86.3 -> 87.0No clear effect
PoP-iter1 vs CSR (prior preference method)PoP LLaVA-1.5 · MMBenchaccuracy: 65.4 -> 68.7Improved
PoP-iter1 vs SeVaPoP LLaVA-1.5 · MM-Vetaccuracy: 37.2 -> 34.1Hurt
train on peer-generated preference data instead of ground-truth SFT on 900K samplesPoP-Vicuna · 15 vision-language benchmarks (average)average score: 54.0 -> 57.0Improved
add self-evaluation by each panel member (ablation)PoP · 15 vision-language benchmarks (average)average scoreNo clear effect
use absolute scoring instead of relative scoring (ablation)PoP · 15 vision-language benchmarks (average)average scoreImproved
use uncertainty-weighted aggregation instead of mean (ablation)PoP · 15 vision-language benchmarks (average)average scoreNo clear effect
use panel-wide ensemble instead of average single model for zero-shot selectionPoP · Best-of-15 selection at four panel scalesaverage scoreImproved
add more OCR data (0% to 100%) to OCR-Dumb panel member (ablation)PoP · five capability categoriesaverage scoreImproved

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