List your company
ResearchFormerly WizWand

Latency Measurement on OOL-Pairs for Long-context reasoning

5.1Latency (s)Direct
-9.30487.923185.15282.377Mar 20, 2026
Updated 7mo ago

Evaluation results

MethodLinks
DirectParadigm=P1: Direct, Backbone=Mistral, Model Size=7B2026.035.1
DirectParadigm=P1: Direct, Backbone=Llama, Model Size=8B2026.035.8
DirectParadigm=P1: Direct, Backbone=Qwen3, Model Size=8B2026.036.2
DirectParadigm=P1: Direct, Backbone=Mistral, Model Size=Cdsrl2026.039.4
DirectParadigm=P1: Direct, Backbone=Qwen3, Model Size=32B2026.0310.8
DirectParadigm=P1: Direct, Backbone=Mistral, Model Size=8x22B2026.0311.7
DirectParadigm=P1: Direct, Backbone=Llama, Model Size=70B2026.0312.3
DirectParadigm=P1: Direct, Backbone=Qwen3, Model Size=235B2026.0328.4
λ-RLMParadigm=P3: λ-RLM, Backbone=Llama, Model Size=30.82026.0330.8
DirectParadigm=P1: Direct, Backbone=Llama, Model Size=405B2026.0332.1
λ-RLMParadigm=P3: λ-RLM, Backbone=Qwen3, Model Size=235B2026.0334.7
λ-RLMParadigm=P3: λ-RLM, Backbone=Mistral, Model Size=Cdsrl2026.0338.2
λ-RLMParadigm=P3: λ-RLM, Backbone=Llama, Model Size=70B2026.0338.4
λ-RLMParadigm=P3: λ-RLM, Backbone=Qwen3, Model Size=32B2026.0341.2
λ-RLMParadigm=P3: λ-RLM, Backbone=Mistral, Model Size=8x22B2026.0344.6
λ-RLMParadigm=P3: λ-RLM, Backbone=Qwen3, Model Size=8B2026.0348.6
λ-RLMParadigm=P3: λ-RLM, Backbone=Llama, Model Size=8B2026.0352.1
λ-RLMParadigm=P3: λ-RLM, Backbone=Mistral, Model Size=7B2026.0355.8
RLMParadigm=P2: RLM, Backbone=Llama, Model Size=405B2026.03156.4
RLMParadigm=P2: RLM, Backbone=Qwen3, Model Size=235B2026.03178.3
RLMParadigm=P2: RLM, Backbone=Mistral, Model Size=Cdsrl2026.03224.5
RLMParadigm=P2: RLM, Backbone=Llama, Model Size=70B2026.03241.8
RLMParadigm=P2: RLM, Backbone=Qwen3, Model Size=32B2026.03264.7
RLMParadigm=P2: RLM, Backbone=Mistral, Model Size=8x22B2026.03288.6
RLMParadigm=P2: RLM, Backbone=Qwen3, Model Size=8B2026.03312.5
RLMParadigm=P2: RLM, Backbone=Llama, Model Size=8B2026.03338.1
RLMParadigm=P2: RLM, Backbone=Mistral, Model Size=7B2026.03365.2