Every Step Counts: Step-Level Credit Assignment for Tool-Integrated Text-to-SQL
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2026
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866917466431553536 |
|---|---|
| author | Dai, Yaxun Sun, Baolin Wang, Junying Wang, Pengfei Gao, Yingqi Dong, Xuemei Chu, Mengdie Qi, Xiang Chao, Pingfu |
| author_facet | Dai, Yaxun Sun, Baolin Wang, Junying Wang, Pengfei Gao, Yingqi Dong, Xuemei Chu, Mengdie Qi, Xiang Chao, Pingfu |
| contents | Tool-integrated Text-to-SQL parsing has emerged as a promising paradigm, framing SQL generation as a sequential decision-making process interleaved with tool execution. However, existing reinforcement learning approaches mainly rely on coarse-grained outcome supervision, resulting in a fundamental credit assignment problem: models receive the same reward for any trajectory that yields the correct answer, even when intermediate steps are redundant, inefficient, or erroneous. Consequently, models are encouraged to explore suboptimal reasoning spaces, limiting both efficiency and generalization. To address this problem, we propose FineStep, a novel framework for step-level credit assignment in tool-augmented Text-to-SQL. First, we introduce a reward design with independent process rewards to alleviate the signal sparsity of outcome supervision. Next, we present a step-level credit assignment mechanism to precisely quantify the value of each reasoning step. Finally, we develop a policy optimization method based on step-level advantages for efficient updates. Extensive experiments on BIRD benchmarks show that FineStep achieves state-of-the-art performance and reduces redundant tool interactions, with a 3.25% average EX gain over GRPO at the 4B scale. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2605_04719 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Every Step Counts: Step-Level Credit Assignment for Tool-Integrated Text-to-SQL Dai, Yaxun Sun, Baolin Wang, Junying Wang, Pengfei Gao, Yingqi Dong, Xuemei Chu, Mengdie Qi, Xiang Chao, Pingfu Computation and Language Tool-integrated Text-to-SQL parsing has emerged as a promising paradigm, framing SQL generation as a sequential decision-making process interleaved with tool execution. However, existing reinforcement learning approaches mainly rely on coarse-grained outcome supervision, resulting in a fundamental credit assignment problem: models receive the same reward for any trajectory that yields the correct answer, even when intermediate steps are redundant, inefficient, or erroneous. Consequently, models are encouraged to explore suboptimal reasoning spaces, limiting both efficiency and generalization. To address this problem, we propose FineStep, a novel framework for step-level credit assignment in tool-augmented Text-to-SQL. First, we introduce a reward design with independent process rewards to alleviate the signal sparsity of outcome supervision. Next, we present a step-level credit assignment mechanism to precisely quantify the value of each reasoning step. Finally, we develop a policy optimization method based on step-level advantages for efficient updates. Extensive experiments on BIRD benchmarks show that FineStep achieves state-of-the-art performance and reduces redundant tool interactions, with a 3.25% average EX gain over GRPO at the 4B scale. |
| title | Every Step Counts: Step-Level Credit Assignment for Tool-Integrated Text-to-SQL |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2605.04719 |