This method requires a fine-tuned model created with Seekr’s fine-tuning feature. The method works with models trained after September 22, 2025. Attribution isn’t available for preference or reinforcement fine-tunes, including continued fine-tunes that use either method.
For a model created with continued fine-tuning, results include only examples from the training data of that model’s own training run. Results never include examples from the fine-tunes it builds on.
Retrieve influential fine-tuning data
Endpoint:GET /v1/flow/explain/models/{model_id}/influential-finetuning-data
model_id takes the deployment ID of a deployed fine-tuned model. A fine-tuning job ID or a model name returns a 404. To find the deployment ID, list your deployments and use the id field.chat.completions call, provide it as answer to skip an extra generation:
Reading the result
The response containsresults, a list of influential question-answer pairs, along with the answer used (echoed back if you provided one, otherwise generated internally) and the schema version. Each result carries its id, the source file_id (use it to trace back and edit the training documents), the messages content in Q: <question>\nA: <answer> form, and an influence_level of high, medium, or low. Irrelevant pairs are filtered out and not returned. For the complete response schema, see the endpoint reference above.
Best practices
- Interpreting influence levels:
highmeans the Q/A pair strongly shaped the response;mediummeans moderate impact;lowmeans minimal contribution. Look for recurringhighpairs to understand the training patterns driving a response. - Unexpected results: If unrelated pairs are surfacing with high influence, review and refine your fine-tuning dataset.
- Empty results: The model may simply not have found training pairs relevant to the prompt. This is not an error. A continued fine-tune also returns no results for a prompt that matches only the training data of an earlier round.
Common errors
- TypeError – Raised by the SDK when a required parameter such as
questionis missing or invalid. - 404 Not found – No deployed model matches the provided
model_id. Confirm you used the deployment ID and that the deployment is active.