List your company

Long-context Question Answering on LV-Eval

14.81F1 ScorePANINI
4.60767.25639.90512.5537Feb 16, 2026

Evaluation results

MethodLinks
PANINIRetrieval Category=Proposed Method2026.0214.81————————————
HippoRAG 2Retrieval Category=Structure-Augmented RAG2026.0214————————————
Qwen3-Embedding (8B)Retrieval Category=Dense Retrieval, Reranker=Qwen3-Embedding (8B) + reranker2026.0211.8————————————
Qwen3-Embedding (8B)Retrieval Category=Dense Retrieval, Reranker=None2026.0211.7————————————
Search-R1Retrieval Category=Agentic Systems2026.0211.5————————————
GraphRAGRetrieval Category=Structure-Augmented RAG2026.0211————————————
NV-Embed-v2 (7B)Retrieval Category=Dense Retrieval2026.0210————————————
RAPTORRetrieval Category=Structure-Augmented RAG2026.029.2————————————
IRCoTRetrieval Category=Agentic Systems2026.029.1————————————
BM25Retrieval Category=Sparse Retrieval, Reranker=BM25 + reranker2026.028.1————————————
BM25Retrieval Category=Sparse Retrieval, Reranker=None2026.027.8————————————
HippoRAGRetrieval Category=Structure-Augmented RAG2026.027.6————————————
NoneRetrieval Category=No Retrieval2026.025————————————
LightRAGRetrieval Category=Structure-Augmented RAG2026.025————————————
Qwen3-0.6BAttention=Full attention, Avg. Len.=128k2026.02—6.3111.181.674.493.196.445.241.424.875.536.445.16
Qwen3-0.6B-ALLMEMArchitecture=ALLMEM, Avg. Len.=128k2026.02—4.899.567.54.84.963.643.261.222.343.733.794.52
Qwen3-0.6B-MambaArchitecture=Hybrid Mamba-2, Avg. Len.=128k2026.02—5.919.251.593.73.294.023.691.362.844.423.63.97
Qwen3-0.6B-SWA-SinksAttention=sliding-window, Window Length=8k, Attention Sinks=128, Avg. Len.=128k2026.02—5.459.331.12.912.243.661.70.882.563.063.063.27