List your company

Long-context reasoning on LongReason

86.9ScoreLONGTRACERL-GRPO
72.65276.35180.0583.749May 29, 2026

Evaluation results

MethodLinks
LONGTRACERL-GRPOBackbone=Qwen3-30B-A3B-Thinking-25072026.0586.9
DocQABackbone=Qwen3-30B-A3B-Thinking-25072026.0586.4
LONGTRACERLBackbone=Qwen3-30B-A3B-Thinking-25072026.0585.4
LoongRLBackbone=Qwen3-30B-A3B-Thinking-25072026.0584.9
BaseBackbone=Qwen3-30B-A3B-Thinking-25072026.0584.2
LongRLVRBackbone=Qwen3-30B-A3B-Thinking-25072026.0584.2
LONGTRACERLBackbone=Qwen3-4B-Thinking-25072026.0583.8
LongRLVRBackbone=Qwen3-4B-Thinking-25072026.0580.7
DocQABackbone=Qwen3-4B-Thinking-25072026.0579.9
LoongRLBackbone=Qwen3-4B-Thinking-25072026.0578.7
LONGTRACERL-GRPOBackbone=Qwen3-4B-Thinking-25072026.0578.7
BaseBackbone=Qwen3-4B-Thinking-25072026.0578.5
LONGTRACERL-GRPOBackbone=DeepSeek-R1-0528-Qwen3-8B2026.0575.4
LONGTRACERLBackbone=DeepSeek-R1-0528-Qwen3-8B2026.0575.2
BaseBackbone=DeepSeek-R1-0528-Qwen3-8B2026.0574.1
DocQABackbone=DeepSeek-R1-0528-Qwen3-8B2026.0573.7
LoongRLBackbone=DeepSeek-R1-0528-Qwen3-8B2026.0573.3
LongRLVRBackbone=DeepSeek-R1-0528-Qwen3-8B2026.0573.2