Multi-Turn Evaluation Benchmarks A collection of benchmarks for evaluating LMs or VLMs under multi-turn interaction passing2961/MultiVerse Viewer • Updated Nov 1, 2025 • 647 • 97 • 1 passing2961/photochat_plus Viewer • Updated Dec 3, 2024 • 968 • 85 • 4 RefineBench/RefineBench Viewer • Updated Dec 2, 2025 • 1k • 1.47k • 5
Thanos Skill-of-Mind-Infused LLM passing2961/Thanos-1B 1B • Updated Nov 8, 2024 • 15 passing2961/Thanos-3B 3B • Updated Nov 8, 2024 • 3 • 4 passing2961/Thanos-8B 8B • Updated Nov 8, 2024 • 3 • 3 passing2961/multifaceted-skill-of-mind Viewer • Updated Nov 8, 2024 • 100k • 40 • 5
Multi-Turn Evaluation Benchmarks A collection of benchmarks for evaluating LMs or VLMs under multi-turn interaction passing2961/MultiVerse Viewer • Updated Nov 1, 2025 • 647 • 97 • 1 passing2961/photochat_plus Viewer • Updated Dec 3, 2024 • 968 • 85 • 4 RefineBench/RefineBench Viewer • Updated Dec 2, 2025 • 1k • 1.47k • 5
Thanos Skill-of-Mind-Infused LLM passing2961/Thanos-1B 1B • Updated Nov 8, 2024 • 15 passing2961/Thanos-3B 3B • Updated Nov 8, 2024 • 3 • 4 passing2961/Thanos-8B 8B • Updated Nov 8, 2024 • 3 • 3 passing2961/multifaceted-skill-of-mind Viewer • Updated Nov 8, 2024 • 100k • 40 • 5
passing2961/finch_4b_kto_held_out_expr_purpose_qwen_max8192_kto_5.0e-7_1.0_train42_cosine Image-Text-to-Text • 5B • Updated about 5 hours ago
passing2961/finch_8b_soft_with_held_out_expr_purpose_qwen_1.0e-5_1.0_train42_cosine Text Generation • 8B • Updated about 7 hours ago
passing2961/finch_4b_soft_without_held_out_expr_purpose_qwen_1.0e-5_1.0_train42_cosine Image-Text-to-Text • 5B • Updated about 7 hours ago
passing2961/finch_4b_hard_without_held_out_expr_purpose_qwen_1.0e-5_1.0_train42_cosine Image-Text-to-Text • 5B • Updated about 10 hours ago
passing2961/finch_8b_soft_without_held_out_expr_purpose_qwen_1.0e-5_1.0_train42_cosine Text Generation • 8B • Updated about 11 hours ago
passing2961/finch_8b_hard_with_held_out_expr_purpose_qwen_1.0e-5_1.0_train42_cosine Text Generation • 8B • Updated about 11 hours ago
passing2961/finch_9b_soft_without_held_out_expr_purpose_qwen_1.0e-5_1.0_train42_cosine Image-Text-to-Text • 9B • Updated about 11 hours ago
passing2961/finch_8b_hard_without_held_out_expr_purpose_qwen_1.0e-5_1.0_train42_cosine Text Generation • 8B • Updated about 19 hours ago
passing2961/finch_2b_soft_without_held_out_expr_purpose_qwen_1.0e-5_1.0_train42_cosine Image-Text-to-Text • 3B • Updated about 19 hours ago
passing2961/finch_9b_hard_without_held_out_expr_purpose_qwen_1.0e-5_1.0_train42_cosine Image-Text-to-Text • 9B • Updated about 20 hours ago