Dialogue Memory Retrieval on LMEB benchmark leaderboard · SOTA2 Research

Dialogue Memory Retrieval on LMEB

66.42Avg. nDCG@10Qwen3-Reranker-8B

Evaluation Results

Method	Links
Qwen3-Reranker-8B2026.06		66.42	64.64	65.91	78.93	76.31	57.13	55.58
mxbai-rerank-large-v22026.06		66.08	70.41	61.33	87.16	81.18	54.43	41.96
KaLM-Reranker-V1-L2026.06		64.16	68.83	59.63	89.73	73.29	55.15	38.35
Qwen3-Reranker-0.6B2026.06		63.69	67.36	60.32	84.91	75.67	53.26	40.63
bge-reranker-v2-gemma2026.06		63.5	67.12	61.3	87.23	76.5	54.74	34.13
KaLM-Reranker-V1-S2026.06		63.15	65.81	59.86	87.58	74.73	53.49	37.44
Qwen3-Reranker-4B2026.06		62.86	66.68	63.43	53.72	78.65	55.91	58.79
mxbai-rerank-base-v22026.06		62.47	69.41	55.76	86.29	79.97	49.75	33.61
bge-reranker-v2-m32026.06		62.38	69.15	60.2	82.39	76.82	53.9	31.84
KaLM-Reranker-V1-N2026.06		61.39	65.11	57.08	85.52	74.83	52.7	33.1
bge-reranker-large2026.06		60.64	62.55	58.94	84.1	71.48	51.5	35.3
jina-reranker-v22026.06		59.6	54.7	56.22	86.55	72.66	51.53	35.97
gte-reranker-base2026.06		56.08	54.78	51.01	78.94	68.76	47.97	35.02
KaLM-Embedding-V2.52026.06		50.8	62.74	41.88	75.18	69.59	38.6	16.82