Text Classification on AG News

Metric: Error (lower is better)

LeaderboardDataset

Loading chart...

Results

#	Model↕	Error▲	Extra Data	Paper	Date↕	Code
1	XLNet	4.45	No	XLNet: Generalized Autoregressive Pretraining fo...	2019-06-19	Code
2	BERT-ITPT-FiT	4.8	No	How to Fine-Tune BERT for Text Classification?	2019-05-14	Code
3	L MIXED	4.95	No	Revisiting LSTM Networks for Semi-Supervised Tex...	2020-09-08	Code
4	ULMFiT	5.01	No	Universal Language Model Fine-tuning for Text Cl...	2018-01-18	Code
5	DRNN	5.53	No	-	-	-
6	DNC+CUW	6.1	No	Learning to Remember More with Less Memorization	2019-01-05	Code
7	ULMFiT (Small data)	6.3	No	Sampling Bias in Deep Active Classification: An ...	2019-09-20	Code
8	CNN	6.57	No	Supervised and Semi-Supervised Text Categorizati...	2016-02-07	-
9	DPCNN	6.87	No	-	-	Code
10	EXAM	7	No	Explicit Interaction Model towards Text Classifi...	2018-11-23	Code
11	SWEM-concat	7.34	No	Baseline Needs More Love: On Simple Word-Embeddi...	2018-05-24	Code
12	Capsule-B	7.4	No	Investigating Capsule Networks with Dynamic Rout...	2018-03-29	Code
13	fastText	7.5	No	Bag of Tricks for Efficient Text Classification	2016-07-06	Code
14	LEAM	7.55	No	Joint Embedding of Words and Labels for Text Cla...	2018-05-10	Code
15	CCCapsNet	7.61	No	Compositional Coding Capsule Network with K-Mean...	2018-10-22	Code
16	Balanced+bi-leaf-RNN	7.9	No	On Tree-Based Neural Sentence Modeling	2018-08-29	Code
17	VDCN	8.67	No	Very Deep Convolutional Networks for Text Classi...	2016-06-06	Code
18	SVDCNN	9.45	No	Squeezed Very Deep Convolutional Neural Networks...	2019-01-28	Code
19	Char-level CNN	9.51	No	Character-level Convolutional Networks for Text ...	2015-09-04	Code
20	Seq2CNN with GWS(50)	9.64	No	Abstractive Text Classification Using Sequence-t...	2018-05-20	Code
21	Llama-3.3-70B + CAPO	11.2	No	CAPO: Cost-Aware Prompt Optimization	2025-04-22	Code
22	Qwen2.5-32B + CAPO	12.93	No	CAPO: Cost-Aware Prompt Optimization	2025-04-22	Code
23	ToWE-SG	14	No	-	-	Code
24	Mistral-Small-24B + CAPO	15.7	No	CAPO: Cost-Aware Prompt Optimization	2025-04-22	Code