Instruction Following Evaluation on InfoBench

86.1Score

RM-R1-32B (Qwen-2.5-Inst)

Updated 5d ago

Evaluation Results

Method	Links
RM-R1-32B (Qwen-2.5-Inst) 2026.05		86.1
Gemini-2.5-Flash 2026.05		85.6
RM-R1-14B (Qwen-2.5-Inst) 2026.05		85.5
RM-R1-32B (DeepSeek-Dist) 2026.05		85
RRM-32B 2026.05		84.4
Rubric-ARROW-voting@5 2026.05		84
Rubric-RM-voting@5 2026.05		83.8
RM-R1-14B (DeepSeek-Dist) 2026.05		82.4
API Prompting (Rubric+Judge pairwise) 2026.05		82.2
Rubric-ARROW 2026.05		82.1
Rubric-RM 2026.05		80.8
RIFL-voting@5 2026.05		80.1
Rubric-ARROW w/o RL-voting@5 2026.05		79.7
RIFL 2026.05		78.4
Rubric-ARROW w/o RL 2026.05		75.9
Qwen-3-8B (Rubric+Judge pairwise) 2026.05		74.6
API Prompting (direct Judge) 2026.05		72.9
RM-R1-7B (Qwen-2.5-Inst) 2026.05		71.3
RM-R1-7B (DeepSeek-Dist) 2026.05		70.3
Qwen-3-8B (Rubric+Judge pointwise) 2026.05		69.9
RRM-7B 2026.05		68.2
JudgeLRM-7B 2026.05		62.7
API Prompting (Rubric+Judge pointwise) 2026.05		44.8