fixing ablate_then_add operator with 0 magnitude steering

[chanind]

When doing ablation, zero-magnitude steering vectors cause NaN issues. This PR skips steering if the magnitude is 0. This should be a minor performance boost as well for the 0 magnitude case.

Summary by CodeRabbit

• Bug Fixes

  • Updated the steering-vectors dependency to version 0.12.1, resolving compatibility issues.

• Tests

  • Added a new test for run_persona_cross_steering_experiment, ensuring functionality with specific parameters.

• Refactor

  • Improved handling logic by skipping patching when direction_multiplier is 0.
  • Simplified multipliers list in evaluate_cross_steering function.

[coderabbitai]

Walkthrough

The changes primarily focus on updating dependencies, refining logic, and enhancing test coverage. The pyproject.toml file reflects a version bump for the steering-vectors dependency. Conditional logic was added in repepo/core/hook.py to avoid unnecessary operations when direction_multiplier is zero. The evaluate_cross_steering function in repepo/steering/evaluate_cross_steering.py saw a minor adjustment in its default multipliers list. Lastly, a new test was introduced in tests/experiments/test_persona_generalization.py.

Changes

File	Change Summary
pyproject.toml	Updated the steering-vectors dependency version from 0.11.0 to 0.12.1.
repepo/core/hook.py	Added a conditional check to skip patching when direction_multiplier is 0.
repepo/steering/evaluate_cross_steering.py	Updated the default multipliers list from [0.0] to [0].
tests/experiments/test_persona_generalization.py	Introduced a new test function test_run_persona_cross_steering_experiment to test specific model and tokenizer parameters.

Sequence Diagram(s)

sequenceDiagram
    participant Test as test_run_persona_cross_steering_experiment
    participant Function as run_persona_cross_steering_experiment
    participant Model as GPTNeoXForCausalLM
    participant Tokenizer as Tokenizer

    Test ->> Function: Call with model and tokenizer
    Function ->> Model: Utilize for processing
    Function ->> Tokenizer: Utilize for tokenizing
    Function -->> Test: Return None

Loading

---

Recent review details

Configuration used: .coderabbit.yaml
Review profile: CHILL

Commits

Files that changed from the base of the PR and between c4b73f4 and 233316b.

Files ignored due to path filters (1)

• pdm.lock is excluded by !**/*.lock

Files selected for processing (4)

• pyproject.toml (1 hunks)
• repepo/core/hook.py (1 hunks)
• repepo/steering/evaluate_cross_steering.py (1 hunks)
• tests/experiments/test_persona_generalization.py (2 hunks)

Files skipped from review due to trivial changes (1)

• pyproject.toml

Additional comments not posted (3)

repepo/core/hook.py (1)

45-57: LGTM! The conditional check for direction_multiplier is a good addition.

This change prevents unnecessary operations and potential NaN issues when the multiplier is 0. It also improves performance for zero-magnitude vectors.

repepo/steering/evaluate_cross_steering.py (1)

86-86: LGTM! The change to the multipliers list is consistent.

Changing the multipliers list from [0.0] to [0] ensures consistency in the data type of multipliers. This should not affect the functionality.

tests/experiments/test_persona_generalization.py (1)

206-221: LGTM! The new test function is well-defined.

The test_run_persona_cross_steering_experiment function is straightforward and verifies that the run_persona_cross_steering_experiment function returns a non-None result.

---

Thank you for using CodeRabbit. We offer it for free to the OSS community and would appreciate your support in helping us grow. If you find it useful, would you consider giving us a shout-out on your favorite social media?

Share

• X
• Mastodon
• Reddit
• LinkedIn

Tips

Chat

There are 3 ways to chat with CodeRabbit:

• Review comments: Directly reply to a review comment made by CodeRabbit. Example:
  • I pushed a fix in commit <commit_id>.
  • Generate unit testing code for this file.
  • Open a follow-up GitHub issue for this discussion.
• Files and specific lines of code (under the "Files changed" tab): Tag @coderabbitai in a new review comment at the desired location with your query. Examples:
  • @coderabbitai generate unit testing code for this file.
  • @coderabbitai modularize this function.
• PR comments: Tag @coderabbitai in a new PR comment to ask questions about the PR branch. For the best results, please provide a very specific query, as very limited context is provided in this mode. Examples:
  • @coderabbitai generate interesting stats about this repository and render them as a table.
  • @coderabbitai show all the console.log statements in this repository.
  • @coderabbitai read src/utils.ts and generate unit testing code.
  • @coderabbitai read the files in the src/scheduler package and generate a class diagram using mermaid and a README in the markdown format.
  • @coderabbitai help me debug CodeRabbit configuration file.

Note: Be mindful of the bot's finite context window. It's strongly recommended to break down tasks such as reading entire modules into smaller chunks. For a focused discussion, use review comments to chat about specific files and their changes, instead of using the PR comments.

CodeRabbit Commands (invoked as PR comments)

• @coderabbitai pause to pause the reviews on a PR.
• @coderabbitai resume to resume the paused reviews.
• @coderabbitai review to trigger an incremental review. This is useful when automatic reviews are disabled for the repository.
• @coderabbitai full review to do a full review from scratch and review all the files again.
• @coderabbitai summary to regenerate the summary of the PR.
• @coderabbitai resolve resolve all the CodeRabbit review comments.
• @coderabbitai configuration to show the current CodeRabbit configuration for the repository.
• @coderabbitai help to get help.

Additionally, you can add @coderabbitai ignore anywhere in the PR description to prevent this PR from being reviewed.

CodeRabbit Configration File (.coderabbit.yaml)

• You can programmatically configure CodeRabbit by adding a .coderabbit.yaml file to the root of your repository.
• Please see the configuration documentation for more information.
• If your editor has YAML language server enabled, you can add the path at the top of this file to enable auto-completion and validation: # yaml-language-server: $schema=https://coderabbit.ai/integrations/schema.v2.json

Documentation and Community

• Visit our Documentation for detailed information on how to use CodeRabbit.
• Join our Discord Community to get help, request features, and share feedback.
• Follow us on X/Twitter for updates and announcements.