Method Drift›Mixture-of-experts routing
ODP
Mixture-of-experts routing
superseded — cited as a baseline and beaten by newer methods
0 papers critique it · 2 beat it on benchmarks
Beaten on benchmarks
Head-to-head results where a newer method reports beating ODP. Values are copied from the source paper's tables — verify against the cited paper.
- EAC-MoE: Expert-Selection Aware Compressor for Mixture-of-Experts Large Language Models
PESF beats ODP · Avg [Mixtral-8x7B, α=0.3]
72.19 vs 71.98
- EAC-MoE: Expert-Selection Aware Compressor for Mixture-of-Experts Large Language Models
PESF beats ODP · Avg [Phi3.5-moe, α=0.3]
69.27 vs 68.92
- EAC-MoE: Expert-Selection Aware Compressor for Mixture-of-Experts Large Language Models
PESF beats ODP · Avg [Deepseek-moe-16b-base, α=0.3]
61.28 vs 61.19
- EAC-MoE: Expert-Selection Aware Compressor for Mixture-of-Experts Large Language Models
PESF beats ODP · Avg [Qwen1.5-MoE-A2.7B, α=0.3]
64.64 vs 64.48
- Ban&Pick: Ehancing Performance and Efficiency of MoE-LLMs via Smarter Routing
Ban beats ODP · Speedup [Deepseek-V2-Lite-Chat]
1.22 vs 1.19
- Ban&Pick: Ehancing Performance and Efficiency of MoE-LLMs via Smarter Routing
Ban beats ODP · GSM8K/AIME24 [Deepseek-V2-Lite-Chat]
66.11 vs 64.76
- Ban&Pick: Ehancing Performance and Efficiency of MoE-LLMs via Smarter Routing
Ban beats ODP · Math500 [Deepseek-V2-Lite-Chat]
25.00 vs 23.20
- Ban&Pick: Ehancing Performance and Efficiency of MoE-LLMs via Smarter Routing
Ban beats ODP · GPQA [Deepseek-V2-Lite-Chat]
24.24 vs 23.23
- Ban&Pick: Ehancing Performance and Efficiency of MoE-LLMs via Smarter Routing
Ban beats ODP · Livecodebench [Deepseek-V2-Lite-Chat]
8.50 vs 7.75
- Ban&Pick: Ehancing Performance and Efficiency of MoE-LLMs via Smarter Routing
Ban beats ODP · HumanEval+ [Deepseek-V2-Lite-Chat]
48.17 vs 46.95
- Ban&Pick: Ehancing Performance and Efficiency of MoE-LLMs via Smarter Routing
Ban beats ODP · Speedup [Deepseek-v2.5-1210]
1.24 vs 1.20
- Ban&Pick: Ehancing Performance and Efficiency of MoE-LLMs via Smarter Routing
Ban beats ODP · Math500 [Deepseek-v2.5-1210]
80.60 vs 80.20
Newer alternatives
Recent methods in the same sub-problem, not yet superseded in the knowledge base.
- May 22, 2026
- May 21, 2026
- KBVQ-MoEKBVQ-MoE: KLT-guided SVD with Bias-Corrected Vector Quantization for MoE Large Language ModelsJan 30, 2026
- Oct 13, 2025