reference

`gold/scifact_200` gold

Promoted as gold: the yardstick a decision rests on.
Promoted from person the dataset's authors, 2026-09-28: promoted before labels existed (0.6.0): its tier was given when it was made

01what it is

tier: gold
dataset: scifact_200
promoted from: person the dataset's authors
why: promoted before labels existed (0.6.0): its tier was given when it was made
items: 200
created: 2026-09-28T14:59:25Z

02scored against it · 26

scored against silverscored by a judgewrote the reference it is scored againstscored against a label set, not a reference setunmarked rows are scored against gold

arm	metric	mean	items
sf_glm_s_e5first_n200_v1 · pair	ndcg_10	0.789	200	inspect
sf_qwen_s_e5first_n200_v1 · pair	ndcg_10	0.771	200	inspect
sf_glm_s_n200_v1	ndcg_10	0.744	200	inspect
sf_glm_s_bm25first_n200_v1 · pair	ndcg_10	0.742	200	inspect
sf_glm_m_n200_v1	ndcg_10	0.742	200	inspect
sf_deepseek_m_n200_v1	ndcg_10	0.740	200	inspect
sf_deepseek_s_n200_v1	ndcg_10	0.738	200	inspect
sf_gemma_m_n200_v1	ndcg_10	0.736	200	inspect
sf_gemma_l_n200_v1	ndcg_10	0.736	200	inspect
sf_llama_l_n200_v1	ndcg_10	0.728	200	inspect
sf_e5_base_n200_v1	ndcg_10	0.719	200	inspect
sf_qwen_s_n200_v1 · repeats	ndcg_10	0.719	200	inspect
sf_qwen_s_n200_v1	ndcg_10	0.718	200	inspect
sf_llama_s_n200_v1	ndcg_10	0.717	200	inspect
sf_llama_l_n200_v1 · repeats	ndcg_10	0.715	200	inspect
sf_qwen_s_bm25first_n200_v1 · pair	ndcg_10	0.714	200	inspect
sf_llama_s_n200_v1 · repeats	ndcg_10	0.713	200	inspect
sf_bge_small_n200_v1	ndcg_10	0.710	200	inspect
sf_llama_m_n200_v1	ndcg_10	0.709	200	inspect
sf_mistral_s_n200_v1	ndcg_10	0.707	200	inspect
sf_gemma_s_n200_v1	ndcg_10	0.702	200	inspect
sf_mpnet_n200_v1	ndcg_10	0.664	200	inspect
sf_minilm_n200_v1	ndcg_10	0.653	200	inspect
sf_bm25_n200_v1	ndcg_10	0.645	200	inspect
sf_random_n200_v1	ndcg_10	0.001	200	inspect
sf_first_k_n200_v1	ndcg_10	0.000	200	inspect

gold/scifact_200 gold

`gold/scifact_200` gold