Instruction Following Evaluation on IFBench

73.2Score

Rubric-ARROW-voting@5

Updated 5d ago

Evaluation Results

Method	Links
Rubric-ARROW-voting@5 2026.05		73.2
RIFL-voting@5 2026.05		69.8
Gemini-2.5-Flash 2026.05		69.3
RIFL 2026.05		69.2
Rubric-ARROW w/o RL-voting@5 2026.05		68.5
Rubric-ARROW 2026.05		68.2
Rubric-RM-voting@5 2026.05		67.1
API Prompting (Rubric+Judge pairwise) 2026.05		66.2
Rubric-ARROW w/o RL 2026.05		65.5
Rubric-RM 2026.05		65.4
RM-R1-14B (Qwen-2.5-Inst) 2026.05		60.8
RRM-32B 2026.05		60.8
API Prompting (direct Judge) 2026.05		60.4
RM-R1-32B (Qwen-2.5-Inst) 2026.05		60.4
RM-R1-14B (DeepSeek-Dist) 2026.05		59
RM-R1-32B (DeepSeek-Dist) 2026.05		58.6
RM-R1-7B (DeepSeek-Dist) 2026.05		56.5
Qwen-3-8B (Rubric+Judge pairwise) 2026.05		55.6
RM-R1-7B (Qwen-2.5-Inst) 2026.05		55.2
RRM-7B 2026.05		53.2
Qwen-3-8B (Rubric+Judge pointwise) 2026.05		50.2
JudgeLRM-7B 2026.05		47.5
API Prompting (Rubric+Judge pointwise) 2026.05		33.6