Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2603.16158 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866912970737451008 |
|---|---|
| author | Kumar, Abhijit Kumar, Natalya Gupta, Shikhar |
| author_facet | Kumar, Abhijit Kumar, Natalya Gupta, Shikhar |
| contents | Critic-free reinforcement learning with verifiable rewards (RLVR) improves code generation by optimizing unit-test pass rates, but GRPO-style updates suffer from coarse credit assignment: a single outcome signal is spread uniformly across long programs even when failure stems from a localized semantic error. We propose Execution-Grounded Credit Assignment (EGCA), which localizes GRPO updates using execution traces. For programs that satisfy algorithmic constraints but fail tests, EGCA executes the candidate and a canonical reference solution (curated once offline; used for analysis, not supervision) under identical instrumentation, identifies the earliest semantic divergence, and assigns advantage only to the corresponding token span while masking downstream tokens. EGCA is a drop-in modification requiring no critic, auxiliary loss, or learned verifier, yielding 82.1% pass@1 on HumanEval (+3.1 over GRPO) and 68.9% on MBPP (+1.5) with 18% wall-clock overhead. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_16158 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Execution-Grounded Credit Assignment for GRPO in Code Generation Kumar, Abhijit Kumar, Natalya Gupta, Shikhar Machine Learning Critic-free reinforcement learning with verifiable rewards (RLVR) improves code generation by optimizing unit-test pass rates, but GRPO-style updates suffer from coarse credit assignment: a single outcome signal is spread uniformly across long programs even when failure stems from a localized semantic error. We propose Execution-Grounded Credit Assignment (EGCA), which localizes GRPO updates using execution traces. For programs that satisfy algorithmic constraints but fail tests, EGCA executes the candidate and a canonical reference solution (curated once offline; used for analysis, not supervision) under identical instrumentation, identifies the earliest semantic divergence, and assigns advantage only to the corresponding token span while masking downstream tokens. EGCA is a drop-in modification requiring no critic, auxiliary loss, or learned verifier, yielding 82.1% pass@1 on HumanEval (+3.1 over GRPO) and 68.9% on MBPP (+1.5) with 18% wall-clock overhead. |
| title | Execution-Grounded Credit Assignment for GRPO in Code Generation |
| topic | Machine Learning |
| url | https://arxiv.org/abs/2603.16158 |