Consistency Meets Verification: Enhancing Test Generation Quality in Large Language Models Without Ground-Truth Solutions
1 st Hamed Taherkhani, 2 nd Alireza DaghighFarsoodeh, 3 rd Mohammad Chowdhury, 4 th Hung Viet Pham, 5 th Hadi Hemmati · 2026
Read Consistency Meets Verification: Enhancing Test Generation Quality in Large Language Models Without Ground-Truth Solutions by 1 st Hamed Taherkhani, 2 nd Alireza DaghighFarsoodeh, 3 rd Mohammad Chowdhury, 4 th Hung Viet Pham, 5 th Hadi Hemmati (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2602-10522.txt.
Description
Hamed Taherkhani and four co-authors improve automated test making without ground-truth code in a 2026 paper. Current methods often verify tests against known implementations, which risks propagating bugs and fails in test-driven development, where the code does not exist yet. The paper asks how to trust tests synthesized from scratch.
Their pipeline, ConVerTest, combines three plans. Self-consistency voting aggregates generated tests. Chain-of-checks refines code, and a dual execution agreement cross-checks code and tests through consensus. Tests on two benchmarks show gains in test validity, line coverage, and mutation scores over baselines. The work aims at autonomous testing agents whose tests can be trusted before any build exists.
Catalog record
| CREATOR | 1 st Hamed Taherkhani, 2 nd Alireza DaghighFarsoodeh, 3 rd Mohammad Chowdhury, 4 th Hung Viet Pham, 5 th Hadi Hemmati |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 54 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2602-10522