公開データセットは
@thedrcat
氏が公開していたデータセットv2,v3,v4(それぞれは前のバージョンを包含している)を指す
Ghostbuster
LLMによるゴーストライターの検知
1st place short solution summary
▲8
2nd place solution with code and data
(pub17th)
3rd place solution
(pub15)
[4th Place Solution] A Summary of Combined Arms Approach
5th place solution - 1.7 million training examples + domain adaptation
6nd place solution with code
▲2989
[7th Place Solution] Generate Data with Non-Instruction-Tuned Models
(▲2350
[8th LB Solution] Linguistic Features: PPL & GLTR(▲1023)
12th place solution: DeBERTa + TF-IDF
(pub168)
13th place solution - Transformers only
(pub41)
14th Place Solution
(117
18th place solution
(39
19th place solution
[21th Solution] Secret Sauce [0.986 Public - Selected Private: 0.932 Best:0.957]
(pub2)
A kaggle newbie's 🥈23rd solution: Just follow-up excellent public works & ideas
24th solution: DeBERTa & TF-IDF Vectorizer Ensemble
25th Place Solution for DAIGT (Public LB: 0.966 Private LB: 0.927)