List your company

Closed-set Reasoning on ARC Challenge (test)

87.3AccuracyMASS-RAG-Qwen3 8B
19.1836.86554.5572.235Apr 20, 2026

Evaluation results

MethodLinks
MASS-RAG-Qwen3 8BRetrieval Strategy=Proposed2026.0487.3
Qwen3 8BRetrieval Strategy=Training-free retrieval2026.0486.4
MASS-RAG-Llama3 8BRetrieval Strategy=Proposed2026.0478.7
MASS-RAG-Mistral 7BRetrieval Strategy=Proposed2026.0472.4
MASS-RAG-Llama2 7BRetrieval Strategy=Proposed2026.0472.2
Self-RAG 7BRetrieval Strategy=Training-based retrieval2026.0467.3
Llama2-FT 7BRetrieval Strategy=Training-based retrieval2026.0465.8
MAIN-RAG-Llama3 8BRetrieval Strategy=Training-free retrieval2026.0461.9
MAIN-RAG-Mistral 7BRetrieval Strategy=Training-free retrieval2026.0458.9
Llama3 8BRetrieval Strategy=Without retrieval2026.0458.8
Mistral 7BRetrieval Strategy=Training-free retrieval2026.0457.1
Llama3 8BRetrieval Strategy=Training-free retrieval2026.0455.6
Mistral 7BRetrieval Strategy=Without retrieval2026.0455.5
Llama2 7BRetrieval Strategy=Training-free retrieval2026.0451
Llama2 13BRetrieval Strategy=Without retrieval2026.0429.4
Llama2 13BRetrieval Strategy=Training-free retrieval2026.0426
Llama2 7BRetrieval Strategy=Without retrieval2026.0421.8