Benchmarks

Compare state-of-the-art methods across the tasks and datasets used to measure AI progress.

309,131 indexed

Task	Dataset	Domain	Results	Last Update
Retrieval-Augmented Question Answering	NQ	retrieval search ragnlp	45	May 20, 2026
Backdoor Attack Defense	Backdoor Attacks (test)	adversarial safety security	45	May 20, 2026
Long-context Question Answering	LongBench N=162	llm chat instructionnlp	45	May 19, 2026
Topic Modeling	CIFAR100, Food101, SUN397 Averaged (test)	nlp	45	May 19, 2026
Topic Modeling	Five text datasets (News-20K, IMDB, Yelp, DailyMail, Twitter)	nlp