# Socratic Conversation Generation on Socratic Debugging of Student Code 227 samples

Source: [Socratic Conversation Generation on Socratic Debugging of Student Code 227 samples](<https://www.sota2.com/research/sota/socratic-conversation-generation-on-socratic-debugging-of-student-code-227-samples>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-05-12T13:35:34.502Z

Evaluation of models performing Socratic Conversation Generation on the Socratic Debugging of Student Code dataset (227 samples), assessed using metrics like % Valid Convs and % Grounded Turns.

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

| Scope | Recorded value |
| --- | --- |
| Task | Socratic Conversation Generation |
| Dataset (including any stated split) | Socratic Debugging of Student Code 227 samples |
| Metric | Valid Conversions Rate |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 14 of 14 |
| Coverage | All results currently returned for this benchmark |

Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol.

## Metrics

- [Valid Conversions Rate](<https://www.sota2.com/research/sota/socratic-conversation-generation-on-socratic-debugging-of-student-code-227-samples.md?metric=Valid%20Conversions%20Rate>): Higher is better; unit: PERCENT.

- [Grounded Turns Rate](<https://www.sota2.com/research/sota/socratic-conversation-generation-on-socratic-debugging-of-student-code-227-samples.md?metric=Grounded%20Turns%20Rate>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Valid Conversions Rate | Result date | Report paper |
| --- | --- | --- | --- |
| GPT-5 | 98.7 | Not recorded | 6909887b3ef7ccd74e5e5b24 |
| GPT-5-mini | 95.9 | Not recorded | 6909887b3ef7ccd74e5e5b24 |
| GPT-5 | 94.8 | Not recorded | 6909887b3ef7ccd74e5e5b24 |
| GPT-5 | 94.3 | Not recorded | 6909887b3ef7ccd74e5e5b24 |
| Claude Sonnet-4.5 | 92.5 | Not recorded | 6909887b3ef7ccd74e5e5b24 |
| GPT-5-mini | 92.1 | Not recorded | 6909887b3ef7ccd74e5e5b24 |
| Claude Sonnet-4.5 | 89 | Not recorded | 6909887b3ef7ccd74e5e5b24 |
| Gemini 2.5-flash | 86.1 | Not recorded | 6909887b3ef7ccd74e5e5b24 |
| Gemini 2.5-flash | 85.8 | Not recorded | 6909887b3ef7ccd74e5e5b24 |
| GPT-5-mini | 85 | Not recorded | 6909887b3ef7ccd74e5e5b24 |
| Claude Haiku-4.5 | 81.5 | Not recorded | 6909887b3ef7ccd74e5e5b24 |
| Gemini 2.5-pro | 78.7 | Not recorded | 6909887b3ef7ccd74e5e5b24 |
| Gemini 2.5-pro | 78.3 | Not recorded | 6909887b3ef7ccd74e5e5b24 |
| Claude Haiku-4.5 | 68.3 | Not recorded | 6909887b3ef7ccd74e5e5b24 |

### GPT-5

Valid Conversions Rate: 98.7. Result ID: 6a032ca247f34ce924e7afbf.

[Reporting paper](<https://www.sota2.com/research/paper/6909887b3ef7ccd74e5e5b24.md>)

Recorded evaluation configuration: {"Effort level":"low-effort","Reasoning":"true"}. This record does not assert independent reproduction.

### GPT-5-mini

Valid Conversions Rate: 95.9. Result ID: 6a032ca247f34ce924e7afc3.

[Reporting paper](<https://www.sota2.com/research/paper/6909887b3ef7ccd74e5e5b24.md>)

Recorded evaluation configuration: {"Effort level":"medium-effort","Reasoning":"true"}. This record does not assert independent reproduction.

### GPT-5

Valid Conversions Rate: 94.8. Result ID: 6a032ca247f34ce924e7afc0.

[Reporting paper](<https://www.sota2.com/research/paper/6909887b3ef7ccd74e5e5b24.md>)

Recorded evaluation configuration: {"Effort level":"medium-effort","Reasoning":"true"}. This record does not assert independent reproduction.

### GPT-5

Valid Conversions Rate: 94.3. Result ID: 6a032ca247f34ce924e7afbe.

[Reporting paper](<https://www.sota2.com/research/paper/6909887b3ef7ccd74e5e5b24.md>)

Recorded evaluation configuration: {"Effort level":"minimal-effort","Reasoning":"true"}. This record does not assert independent reproduction.

### Claude Sonnet-4.5

Valid Conversions Rate: 92.5. Result ID: 6a032ca247f34ce924e7afc5.

[Reporting paper](<https://www.sota2.com/research/paper/6909887b3ef7ccd74e5e5b24.md>)

Recorded evaluation configuration: {"Reasoning":"true"}. This record does not assert independent reproduction.

### GPT-5-mini

Valid Conversions Rate: 92.1. Result ID: 6a032ca247f34ce924e7afc2.

[Reporting paper](<https://www.sota2.com/research/paper/6909887b3ef7ccd74e5e5b24.md>)

Recorded evaluation configuration: {"Effort level":"low-effort","Reasoning":"true"}. This record does not assert independent reproduction.

### Claude Sonnet-4.5

Valid Conversions Rate: 89. Result ID: 6a032ca247f34ce924e7afc4.

[Reporting paper](<https://www.sota2.com/research/paper/6909887b3ef7ccd74e5e5b24.md>)

Recorded evaluation configuration: {"Reasoning":"false"}. This record does not assert independent reproduction.

### Gemini 2.5-flash

Valid Conversions Rate: 86.1. Result ID: 6a032ca247f34ce924e7afc8.

[Reporting paper](<https://www.sota2.com/research/paper/6909887b3ef7ccd74e5e5b24.md>)

Recorded evaluation configuration: {"Reasoning":"false"}. This record does not assert independent reproduction.

### Gemini 2.5-flash

Valid Conversions Rate: 85.8. Result ID: 6a032ca247f34ce924e7afc9.

[Reporting paper](<https://www.sota2.com/research/paper/6909887b3ef7ccd74e5e5b24.md>)

Recorded evaluation configuration: {"Reasoning":"true"}. This record does not assert independent reproduction.

### GPT-5-mini

Valid Conversions Rate: 85. Result ID: 6a032ca247f34ce924e7afc1.

[Reporting paper](<https://www.sota2.com/research/paper/6909887b3ef7ccd74e5e5b24.md>)

Recorded evaluation configuration: {"Effort level":"minimal-effort","Reasoning":"true"}. This record does not assert independent reproduction.

### Claude Haiku-4.5

Valid Conversions Rate: 81.5. Result ID: 6a032ca247f34ce924e7afc7.

[Reporting paper](<https://www.sota2.com/research/paper/6909887b3ef7ccd74e5e5b24.md>)

Recorded evaluation configuration: {"Reasoning":"true"}. This record does not assert independent reproduction.

### Gemini 2.5-pro

Valid Conversions Rate: 78.7. Result ID: 6a032ca247f34ce924e7afcb.

[Reporting paper](<https://www.sota2.com/research/paper/6909887b3ef7ccd74e5e5b24.md>)

Recorded evaluation configuration: {"Reasoning":"true"}. This record does not assert independent reproduction.

### Gemini 2.5-pro

Valid Conversions Rate: 78.3. Result ID: 6a032ca247f34ce924e7afca.

[Reporting paper](<https://www.sota2.com/research/paper/6909887b3ef7ccd74e5e5b24.md>)

Recorded evaluation configuration: {"Reasoning":"false"}. This record does not assert independent reproduction.

### Claude Haiku-4.5

Valid Conversions Rate: 68.3. Result ID: 6a032ca247f34ce924e7afc6.

[Reporting paper](<https://www.sota2.com/research/paper/6909887b3ef7ccd74e5e5b24.md>)

Recorded evaluation configuration: {"Reasoning":"false"}. This record does not assert independent reproduction.

Page 1 of 1. This is an index page, not the full collection.
