martinjosifoski/SynthIE

The paper ``Exploiting Asymmetry for Synthetic Training Data Generation: SynthIE and the Case of Information Extraction'' builds on the idea that even for hard tasks of interest (with input X and Y) -- for which human-annotation is not practical and high-quality annotated data is not available -- by reversing the task (from Y to X), useful data can be synthetically generated even when that original task cannot be solved directly by the LLM. This process enables the creation of a high-quality dataset of X-Y pairs that will enable the training/fine-tuning of models for the original task of inter

<iframe src="https://silytics.si/embed/item/3c71bdf6-a301-41f8-991b-182462c5fd57" width="420" height="190" style="border:0" loading="lazy" title="Silytics entry"></iframe>

Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 18 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.