Add-claude-3.7 to repro journal #222
Conversation
![korbit-ai[bot]](https://avatars.githubusercontent.com/in/322216?s=60&v=4){kind=small}
There was a problem hiding this comment.
Review by Korbit AI
Korbit automatically attempts to detect when you fix issues in new commits.
| Category | Issue | Fix Detected |
|---|---|---|
 |
Invalid Claude Model Version ▹ view | |
 |
Non-intuitive number format ▹ view |
Files scanned
| File Path | Reviewed |
|---|---|
| src/agentlab/agents/generic_agent/init.py | ✅ |
| main.py | ✅ |
| src/agentlab/llm/llm_configs.py | ✅ |
| src/agentlab/agents/generic_agent/agent_configs.py | ✅ |
Explore our documentation to understand the languages and file types we support and the files we ignore.
Need a new review? Comment
/korbit-reviewon this PR and I'll review your latest changes.Korbit Guide: Usage and Customization
Interacting with Korbit
- You can manually ask Korbit to review your PR using the
/korbit-reviewcommand in a comment at the root of your PR.- You can ask Korbit to generate a new PR description using the
/korbit-generate-pr-descriptioncommand in any comment on your PR.- Too many Korbit comments? I can resolve all my comment threads if you use the
/korbit-resolvecommand in any comment on your PR.- Chat with Korbit on issues we post by tagging @korbit-ai in your reply.
- Help train Korbit to improve your reviews by giving a 👍 or 👎 on the comments Korbit posts.
Customizing Korbit
- Check out our docs on how you can make Korbit work best for you and your team.
- Customize Korbit for your organization through the Korbit Console.
Current Korbit Configuration
General Settings
Setting Value Review Schedule Automatic excluding drafts Max Issue Count 10 Automatic PR Descriptions ✅ Issue Categories
Category Enabled Documentation ✅ Logging ✅ Error Handling ✅ Readability ✅ Design ✅ Performance ✅ Security ✅ Functionality ✅ Feedback and Support
Note
Korbit Pro is free for open source projects 🎉
Looking to add Korbit to your team? Get started with a free 2 week trial here
Your free trial of Korbit Pro is ending in 1 day. To continue using Pro, manage your subscription in the Korbit Console. Otherwise your account will be set to Korbit Starter beginning on February 26th, 2025.
| AGENT_37_SONNET = GenericAgentArgs( | ||
| chat_model_args=CHAT_MODEL_ARGS_DICT["openrouter/anthropic/claude-3.7-sonnet"], | ||
| flags=FLAGS_GPT_4o, | ||
| ) |
There was a problem hiding this comment.
Invalid Claude Model Version 
Tell me more
What is the issue?
The code references a non-existent Claude 3.7 model. Claude 3 only has 3.0 (Opus/Sonnet), 3.1, and 3.2 variants.
Why this matters
Using a non-existent model ID will cause runtime failures when the agent attempts to make API calls.
Suggested change ∙ Feature Preview
Correct the model version to use an existing Claude model. For example:
AGENT_3_SONNET = GenericAgentArgs(
chat_model_args=CHAT_MODEL_ARGS_DICT["openrouter/anthropic/claude-3-sonnet"],
flags=FLAGS_GPT_4o,
)
💬 Chat with Korbit by mentioning @korbit-ai.
| max_total_tokens=200_000, | ||
| max_input_tokens=200_000, | ||
| max_new_tokens=8_192, | ||
| temperature=1e-1, |
There was a problem hiding this comment.
Non-intuitive number format 
Tell me more
What is the issue?
Scientific notation 1e-1 is used repeatedly for temperature values, which is less immediately readable than its decimal equivalent 0.1.
Why this matters
Using scientific notation for a simple decimal value adds unnecessary cognitive load when reading the configuration.
Suggested change ∙ Feature Preview
Replace all instances of temperature=1e-1 with temperature=0.1
💬 Chat with Korbit by mentioning @korbit-ai.
Description by Korbit AI
What change is being made?
Add support for the "Claude 3.7 Sonnet" agent to the codebase and update reproducibility settings in the journal.
Why are these changes being made?
The addition of the "Claude 3.7 Sonnet" agent aims to expand the range of agents available for experiments, enhancing flexibility and capability in study configurations. Updating the reproducibility settings ensures assessments can be reliably recreated, which is crucial for consistent experiment results.