Causal Question Answering on Causal Rivers real-world zero-shot
98.82ACCSTReasoner
Evaluation Results
| Method | Links | |
|---|---|---|
| STReasonerPrompting=TS Encoder, Tokens=0.26M, Cost=$0.052026.01 | 98.82 | |
| Claude-4.5-SonnetPrompting=Text, Tokens=0.78M, Cost=$2.352026.01 | 83.18 | |
| GPT-5.2Prompting=Text, Tokens=3.49M, Cost=$6.152026.01 | 22.32 |