Research map / Learning theory and algorithms / Combinatorial and search optimization
Selecting Representative Examples for Program Synthesis
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Select examples by iteratively adding the least anticipated input-output pair, then initialize CEGIS with the selected subset | Program synthesis with CEGIS · DFA synthesis | average synthesis time | Helped |
| Select examples by iteratively adding the least anticipated input-output pair, then initialize CEGIS with the selected subset | Program synthesis with CEGIS · DFA synthesis | stability of total time | Helped |
| Select examples by iteratively adding the least anticipated input-output pair, then initialize CEGIS with the selected subset | Program synthesis with CEGIS · Programmatic drawing synthesis | average synthesis time | Helped |
| Select examples by iteratively adding the least anticipated input-output pair, then initialize CEGIS with the selected subset | Program synthesis with CEGIS · Programmatic drawing synthesis | variance in total time | Helped |
| Use the learned selection procedure to construct a representative subset | Ordering synthesis · Ordering synthesis | representative subsets: 85% | Helped |
| Select examples with rand80 instead of using the learned representative-example procedure | Ordering synthesis · Ordering synthesis | representative subsets: 30% | Hurt |
| Select examples with CEGIS instead of constructing a representative subset upfront | Program synthesis · Ordering synthesis | representative subsets | Hurt |
| Choose counterexamples at random rather than selecting representative examples upfront | Program synthesis with CEGIS · Programmatic drawing synthesis | overall synthesis time | Hurt |
| Choose counterexamples in a fixed but arbitrary order rather than selecting representative examples upfront | Program synthesis with CEGIS · Programmatic drawing synthesis | overall synthesis time | Hurt |
| Use the top-left-most counterexamples selected by standard CEGIS | Program synthesis with CEGIS · Programmatic drawing synthesis | average synthesis time | No clear effect |
Closest papers by meaning
- Guiding Program Synthesis by Learning to Generate Examples (ICLR 2020 · similarity 0.89)
- Program Synthesis via Test-Time Transduction (NeurIPS 2025 · similarity 0.88)
- Online Prompt Selection for Program Synthesis (AAAI 2025 · similarity 0.86)
- Generating Pragmatic Examples to Train Neural Program Synthesizers (ICLR 2024 · similarity 0.85)
- Solver-Aware Decompositions for Programming-by-Example: When Dividing Requires Knowing how to Conquer (NeurIPS 2026 · similarity 0.85)
- Eco Search: A No-delay Best-First Search Algorithm for Program Synthesis (AAAI 2025 · similarity 0.85)
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.