#alta2023 — Public Fediverse posts

We'll shortly be announcing our Shared Task for #ALTA2024.

In the meantime, we are sharing the 3rd-placed winner of the #ALTA2023 Shared Task, which was to distinguish #LLM-generated from human-generated text.

Here, Ahmed El-Sayed and Omar Nasr of the Arab Academy for Science and Technology use an #ensemble approach of many models to perform well on both the dev and test sets.

https://aclanthology.org/2023.alta-1.20/

While you're thinking about what to submit to the Call for Problems for the #ALTA2024 Shared Task (link below), we're sharing with you the 2nd-place winner of the #ALTA2023 Shared Task, where participants distinguished between #LLM-generated and human-generated text.

Here, Yunhao Fang of #UniMelb used #fineTuning and #EnsembleModels to achieve accuracy of 99%

🔗 Call for Problems for Shared Task: https://alta2024.alta.asn.au/calls

🔗 Paper: https://aclanthology.org/2023.alta-1.19.pdf

While you consider submitting to the Call for Problems for the #ALTA2024 Shared Task (see link below), we'd like to share with you the winner of the #ALTA2023 Shared Task, which involved distinguishing #LLM-generated from human-generated text.

Here, Rinaldo Gagiano and Lin Tian from #RMIT use a fine-tuned #Falcon7B model with label smoothing, yielding an accuracy of 99.91%. Well done!

🔗 Call for Problems for Shared Task: https://alta2024.alta.asn.au/calls

🔗 Paper: https://aclanthology.org/2023.alta-1.18/