List your company

Multi-hop Question Answering on MusiQue answerable setting

38.98ConcisenessGemini-3-pro
0.905610.790320.67530.5597Jan 21, 2026

Evaluation results

MethodLinks
Gemini-3-proSize=UNK2026.0138.9859.2194.1
GPT-o3Size=UNK2026.0137.360.8396.48
Qwen3Size=235B-A22B2026.0135.2560.9495.62
DeepSeek-R1Size=671B2026.0134.3355.8291.12
GPT-4oSize=UNK2026.0132.5457.193.66
Qwen3Size=30B-A3B2026.0132.4153.6985.06
Qwen3Size=14B2026.0131.7848.8185.6
Qwen3Size=8B2026.0131.3352.0686.71
DeepSeek-V3.2Size=671B2026.0130.3459.586.33
Qwen3Size=4B2026.0129.8346.8884.4
LLaMA-3.1Size=70B2026.0127.7251.9286.06
GPT-5.1Size=UNK2026.0127.2760.1187.79
Claude-4.5Size=UNK2026.0127.1363.8889.66
Qwen3Size=1.7B2026.0121.5630.3973.05
SelfCiteSize=8B2026.0119.0235.1675.59
LongCiteSize=8B2026.0117.5532.7376.04
LLaMA-3.1Size=8B2026.0112.3528.5565.96
Qwen3Size=0.6B2026.016.9211.3258.12
FRONTSize=13B2026.016.3413.761.19
LLaMA-3.2Size=3B2026.013.467.4355.42
LLaMA-3.2Size=1B2026.012.373.8226.03