AI Paper Atlas

Research map / Architectures and efficiency / Mixture of experts

ACO-MoE-LoRA: Evolving-while-Training for Adapting Segment Anything Model 2 to Specialized Domains

Kaiyi Luo, Bangjun Wang, Li Zhang, Fanzhang Li, Fei Zhu, Jiaqing Fan

ICML 2026 · Architectures and efficiency › Mixture of experts · approach: Mixture-of-experts LoRA

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 encoder adaptation (LoRA) to frozen SAM2 with U-Net decoder (ablation)Baseline · Polyp avgmDice: 0.641 -> 0.858Improved
add latent MoE with random gating (ablation)Encoder LoRA model · Polyp avgmDice: 0.858 -> 0.853No clear effect
use gradient gate instead of random gate in latent MoE (ablation)Latent MoE · Polyp avgmDice: 0.853 -> 0.857Improved
use auxiliary-loss-free gate instead of random gate (ablation)Latent MoE · Polyp avgmDice: 0.853 -> 0.858Improved
add pheromone-guided gating (ablation)Latent MoE · Polyp avgmDice: 0.853 -> 0.865Improved
add ACO-driven elastic rank (ablation)ACO-MoE-LoRA · Polyp avgmDice: 0.865 -> 0.871Improved
use dynamic evolved rank instead of best fixed rank (r=16) (ablation)ACO-MoE-LoRA decoder · Polyp avgmDice: 0.865 -> 0.871Improved
use fixed rank r=4 instead of r=16 (ablation)ACO-MoE-LoRA decoder · Polyp avgmDice: 0.865 -> 0.852Hurt
use fixed rank r=32 instead of r=16 (ablation)ACO-MoE-LoRA decoder · Polyp avgmDice: 0.865 -> 0.859Hurt
use ACO instead of GA as meta-controller (ablation)ACO-MoE-LoRA · PolypmDice: 0.862 -> 0.871Improved
use ACO instead of PSO as meta-controller (ablation)ACO-MoE-LoRA · PolypmDice: 0.868 -> 0.871Improved
turn on pheromone at inference (ablation)ACO-MoE-LoRA · Polyp testmDice: 0.865 -> 0.865No clear effect

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