Research map / Large language models / Agents, code and math
ExPairT-LLM: Exact Learning for LLM Code Selection by Pairwise Queries
Tom Yuviler, Dana Drachsler-Cohen
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add pairwise equivalence queries to the tournament (ablation) | ExPairT-LLM · HumanEval (avg over 5 models) | pass@1: +7.7% | Improved |
| add pairwise membership queries instead of choosing the largest cluster (ablation) | ExPairT-LLM · HumanEval (avg over 5 models) | pass@1: +24.6% | Improved |
| use ExPairT-LLM selection instead of B4 | code selection with 25 candidates · HumanEval/MBPP/APPS, 5 candidate models | pass@1: +13.0% on average | Improved |
| use ExPairT-LLM instead of B4 on MBPP with Codex candidates | code selection · MBPP-sanitized (Codex candidates) | pass@1: 84.4 vs 84.4 | No clear effect |
| apply ExPairT-LLM to o1-mini outputs | OpenAI o1-mini · APPS (150 random tasks) | pass@1: 54.7 -> 91.2 | Improved |
| apply ExPairT-LLM to DeepSeek-R1 outputs | DeepSeek-R1 · LiveCodeBench (511 tasks) | pass@1: 55.7 -> 88.4 | Improved |
| apply ExPairT-LLM to Gemini 2.5 Flash outputs | Gemini 2.5 Flash · APPS (150 random tasks) | pass@1: 54.1 -> 70.3 | Improved |
| validate differentiating inputs by executing both programs | ExPairT-LLM · design (no ablation) | robustness to LLM errors | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- CODET — CodeT: Code Generation with Generated Tests (ICLR 2023) · better on 5 of 5 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- rStar-Coder: Scaling Competitive Code Reasoning with a Large-Scale Verified Dataset (NeurIPS 2025 · similarity 0.88)
- How efficient is LLM-generated code? A rigorous & high-standard benchmark (ICLR 2025 · similarity 0.87)
- Online Prompt Selection for Program Synthesis (AAAI 2025 · similarity 0.87)
- Mercury: A Code Efficiency Benchmark for Code Large Language Models (NeurIPS 2024 · similarity 0.86)
- EffiBench-X: A Multi-Language Benchmark for Measuring Efficiency of LLM-Generated Code (NeurIPS 2025 · similarity 0.86)
- CRAFT: Customizing LLMs by Creating and Retrieving from Specialized Toolsets (ICLR 2024 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
