Conversational Success Rate on BIRD-Interact
55.7Task Success RateDIA
Evaluation Results
| Method | Links | |
|---|---|---|
| DIADialect=PostgreSQL2026.06 | 55.7 | |
| MERITDialect=PostgreSQL, LLM=GPT-5.42026.06 | 22.7 | |
| MERITDialect=PostgreSQL, LLM=Claude Opus 4.62026.06 | 20.8 | |
| Claude Opus 4.6Dialect=PostgreSQL2026.06 | 17.5 |