Uncertainty Quantification on LCB Hard
91.1PRRBayes Belief State
Evaluation results
| Method | Links | |
|---|---|---|
| Bayes Belief StateAgent=SAGE agent (Suri et al., 2025), Base LLM=gpt-oss-20b2026.06 | 91.1 | |
| Seq. Prob.Agent=SAGE agent (Suri et al., 2025), Base LLM=gpt-oss-20b2026.06 | 81.4 | |
| Tool Success RateAgent=SAGE agent (Suri et al., 2025), Base LLM=gpt-oss-20b2026.06 | 74.7 | |
| PerplexityAgent=SAGE agent (Suri et al., 2025), Base LLM=gpt-oss-20b2026.06 | 57.8 |