Logical Reasoning on ProofWriter (Acc, TTFT, Delay (s))
81.3AccuracyStreamingThinker
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| StreamingThinkerParadigm=Streaming, Reasoning Depth=D3, Model=Qwen3-4B2025.10 | 81.3 | 20.51 | 11.05 | |
| InterleavedParadigm=Interleaved, Reasoning Depth=D3, Model=Qwen3-4B2025.10 | 80.1 | 20.51 | 22.35 | |
| StreamingThinkerParadigm=Streaming, Reasoning Depth=D2, Model=Qwen3-4B2025.10 | 76.1 | 20.51 | 6.5 | |
| InterleavedParadigm=Interleaved, Reasoning Depth=D2, Model=Qwen3-4B2025.10 | 75 | 20.51 | 18.07 | |
| BatchParadigm=Batch, Model=Qwen3-4B2025.10 | 62 | 232.11 | 61.99 | |
| StreamingThinkerParadigm=Streaming, Reasoning Depth=D1, Model=Qwen3-4B2025.10 | 58.8 | 20.51 | 0.71 | |
| InterleavedParadigm=Interleaved, Reasoning Depth=D1, Model=Qwen3-4B2025.10 | 58.3 | 20.51 | 11.96 |