Method Drift›Mixture-of-experts routing
MoE-LLaVA
MoE-LLaVA: Mixture of Experts for Large Vision-Language ModelsMixture-of-experts routing · first seen Jan 29, 2024
superseded — cited as a baseline and beaten by newer methods
0 papers critique it · 3 beat it on benchmarks
Beaten on benchmarks
Head-to-head results where a newer method reports beating MoE-LLaVA. Values are copied from the source paper's tables — verify against the cited paper.
- Input Domain Aware MoE: Decoupling Routing Decisions from Task Optimization in Mixture of Experts
IDA-MoE beats MoE-LLaVA · MMVet [StableLM-1.6B + CLIP-336]
29.5 vs 26.9
- Input Domain Aware MoE: Decoupling Routing Decisions from Task Optimization in Mixture of Experts
IDA-MoE beats MoE-LLaVA · MME-per [StableLM-1.6B + CLIP-336]
1372.0 vs 1318.2
- Input Domain Aware MoE: Decoupling Routing Decisions from Task Optimization in Mixture of Experts
IDA-MoE beats MoE-LLaVA · PoPE [StableLM-1.6B + CLIP-336]
86.8 vs 85.7
- Input Domain Aware MoE: Decoupling Routing Decisions from Task Optimization in Mixture of Experts
IDA-MoE beats MoE-LLaVA · SQA^I [StableLM-1.6B + CLIP-336]
65.1 vs 62.6
- Input Domain Aware MoE: Decoupling Routing Decisions from Task Optimization in Mixture of Experts
IDA-MoE beats MoE-LLaVA · TextVQA [StableLM-1.6B + CLIP-336]
51.9 vs 50.1
- Input Domain Aware MoE: Decoupling Routing Decisions from Task Optimization in Mixture of Experts
IDA-MoE beats MoE-LLaVA · VizWiz [StableLM-1.6B + CLIP-336]
43.1 vs 36.2
- Input Domain Aware MoE: Decoupling Routing Decisions from Task Optimization in Mixture of Experts
IDA-MoE beats MoE-LLaVA · MMB [StableLM-1.6B + CLIP-336]
63.2 vs 60.2
- EvoMoE: Expert Evolution in Mixture of Experts for Multimodal Large Language Models
EvoMoE beats MoE-LLaVA · AVG [1-2B Sparse Model (S-1.6B)]
67.0 vs 65.9
Newer alternatives
Recent methods in the same sub-problem, not yet superseded in the knowledge base.