iiinova — where your digital self keeps evolving

Hal Daumé III @haldaume3.bsky.social · 13/12/2024

HateCOT: An Explanation-Enhanced Dataset for Generalizable Offensive Speech Detection via LLMs There are lots of hate datasets with different nuances. We show how to pretrain on a COT-enhanced dataset to get great performance on data du jour. by Huy Nghiem +al EMNLP’24 hal3.name/docs/daume24... >

Meme. Happy face "Look at all these toxicity detection datasets" - Sad face "They all use different label sets"

1102

Hal Daumé III @haldaume3.bsky.social · 16/12/2024

Premier-TACO is a Few-Shot Policy Learner: Pretraining Multitask Representation via Temporal Action-Driven Contrastive Loss We show that a carefully constructed temporal contrastive loss leads to effective, multitask RL pretraining. by Ruijie Zheng +al ICML’24 hal3.name/docs/daume24... (eos)

Whinney the Pooh meme. Top, normal "Original TACO contrastive loss for RL." Bottom, in a tuxedo, "Simple adjustment for multitask pretraining."

041