List your company

Multi-turn Text-to-SQL on CoSQL Out-of-domain

74Execution Accuracy[Long-Horizon] Warm-Start SFT + RL (Outcome Only)
55.07259.98664.969.814Oct 12, 2025

Evaluation results

MethodLinks
[Long-Horizon] Warm-Start SFT + RL (Outcome Only)Model Size=4B, Horizon Setting=Long-Horizon2025.107463
[Long-Horizon] Warm-Start SFT + RL (Outcome + Process)Model Size=4B, Horizon Setting=Long-Horizon2025.107463
[Long-Horizon] + Warm-Start SFT Only (Round 2)Model Size=4B, Horizon Setting=Long-Horizon, Learning Round=Round 22025.1073.561.2
[Long-Horizon] Warm-Start SFT Only (Round 1)Model Size=4B, Horizon Setting=Long-Horizon, Learning Round=Round 12025.1072.758.7
[Long-Horizon] + Warm-Start SFT Only (Round 3)Model Size=4B, Horizon Setting=Long-Horizon, Learning Round=Round 32025.1072.361.8
[Short-Horizon] Direct RLModel Size=4B, Horizon Setting=Short-Horizon2025.1071.461.3
[Long-Horizon] Warm-Start SFT + RL (Outcome + Process)Model Size=1.7B, Horizon Setting=Long-Horizon2025.1070.459.8
[Long-Horizon] Warm-Start SFT + RL (Outcome Only)Model Size=1.7B, Horizon Setting=Long-Horizon2025.1070.359.8
[Short-Horizon] SFTModel Size=4B, Horizon Setting=Short-Horizon2025.1070.261
[Long-Horizon] + Warm-Start SFT Only (Round 3)Model Size=1.7B, Horizon Setting=Long-Horizon, Learning Round=Round 32025.1068.856.2
[Long-Horizon] + Warm-Start SFT Only (Round 2)Model Size=1.7B, Horizon Setting=Long-Horizon, Learning Round=Round 22025.1067.254.2
[Long-Horizon] Warm-Start SFT Only (Round 1)Model Size=1.7B, Horizon Setting=Long-Horizon, Learning Round=Round 12025.1067.155.5
[Short-Horizon] Direct RLModel Size=1.7B, Horizon Setting=Short-Horizon2025.1066.355.9
[Short-Horizon] SFTModel Size=1.7B, Horizon Setting=Short-Horizon2025.1064.155.2
CoE-SQLModel Size=Closed-Source, Prompting Strategy=Few-shot, 16-shot2025.1058.549.6
RASAT+PICARDModel Size=3B2025.1055.848