Improve LLM rank candidate selection logic

[coffee3699]

Description

This PR addresses the recall limitation in DirectoryScanner.rank. Currently, the system only sends the first 20 candidates (fixed prefix) to the LLM for ranking, which introduces significant position bias.

By switching from [:top_k] to random.sample(), we ensure a broader coverage of the directory structure without increasing LLM context costs.

For the detailed motivation behind this PR, please see the related issue.

Related Issue

Fixes # 63

Changes

• Modified DirectoryScanner.rank to use random sampling instead of fixed prefix slicing.
• Added safety check for cases where candidate count is less than top_k.

[gemini-code-assist]

Summary of Changes

Hello, I'm Gemini Code Assist¹! I'm currently reviewing this pull request and will post my feedback shortly. In the meantime, here's a summary to help you and other reviewers quickly get up to speed!

This pull request enhances the candidate selection logic in DirectoryScanner.rank to mitigate position bias and improve recall. It introduces random sampling to ensure a more representative set of candidates is sent to the LLM for ranking, while also adding a safety check for scenarios with fewer candidates than the desired sample size.

Highlights

• Recall Improvement: Addresses the recall limitation in DirectoryScanner.rank by ensuring broader coverage of the directory structure.
• Random Sampling: Switched from fixed prefix slicing ([:top_k]) to random.sample() for candidate selection.
• Safety Check: Added a safety check to handle cases where the candidate count is less than top_k.

🧠 New Feature in Public Preview: You can now enable Memory to help Gemini Code Assist learn from your team's feedback. This makes future code reviews more consistent and personalized to your project's style. Click here to enable Memory in your admin console.

Using Gemini Code Assist

The full guide for Gemini Code Assist can be found on our documentation page, here are some quick tips.

Invoking Gemini

You can request assistance from Gemini at any point by creating a comment using either /gemini <command> or @gemini-code-assist <command>. Below is a summary of the supported commands on the current page.

Feature	Command	Description
Code Review	/gemini review	Performs a code review for the current pull request in its current state.
Pull Request Summary	/gemini summary	Provides a summary of the current pull request in its current state.
Comment	@gemini-code-assist	Responds in comments when explicitly tagged, both in pull request comments and review comments.
Help	/gemini help	Displays a list of available commands.

Customization

To customize Gemini Code Assist for GitHub experience, repository maintainers can create a configuration file and/or provide a custom code review style guide (such as PEP-8 for Python) by creating and adding files to a .gemini/ folder in the base of the repository. Detailed instructions can be found here.

Limitations & Feedback

Gemini Code Assist may make mistakes. Please leave feedback on any instances where its feedback is incorrect or counter productive. You can react with 👍 and 👎 on @gemini-code-assist comments. If you're interested in giving your feedback about your experience with Gemini Code Assist for GitHub and other Google products, sign up here.

Footnotes

1. Review the Privacy Notices, Generative AI Prohibited Use Policy, Terms of Service, and learn how to configure Gemini Code Assist in GitHub here. Gemini can make mistakes, so double check it and use code with caution. ↩

[gemini-code-assist]

Code Review

This pull request improves the candidate selection logic for LLM ranking by introducing random sampling, which helps mitigate position bias. The change from simple slicing to random.sample is a good improvement for recall. My review includes a suggestion to make the new logic more robust by handling edge cases for the top_k parameter to prevent potential runtime errors.