inconsistent results and explanations
Open
Nobody has claimed this yet.
- Dominant language
- C
- Stars
- 19
- Forks
- 1
- PR merge metrics
- No merged PRs in 30d
Description
Some benchmarks show no improvement (e.g., GPT-4o on Prefix Sum)
Performance on NAS benchmarks is minimal
Achieve more consistent improvements
- by running the LLM multiple times and using an ensemble or majority vote to reduce randomness, or
- by fine-tuning an open-source model on a small set of compiler-report-to-optimized-code examples to guide it.
explain why certain benchmarks see dramatic improvements while others don't
Contributor guide
No contributing guide indexed for this repository
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
The issue names no files, tests, or entry points. Start by locating the benchmark and model-evaluation code, then clarify which benchmarks, consistency measure, and explanation scope are intended. Done should be defined by agreed evaluation results and explanations for differing benchmark outcomes.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- c
- Domain
- compilers, machine-learning
- Issue type
- Feature
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 25/100