Share your thoughts, 1 month free Claude Pro on usSee more

Knowledge Retention

Benchmarks

Dataset Name	SOTA Method	Metric
MMLU (full)		MMLU Accuracy88.3	60	1mo ago
Knowledge Retention Evaluation Set	SaLoRA	Cosine Similarity98.04	49	27d ago
WMDP retain	NPO	Retain48.3	36	19d ago
Source domain	LaLoRA	Source Domain Accuracy65.9	23	3mo ago
RWKU Neighbor Set	DirectQA	FB Score64.64	17	1mo ago
WMDP cyber (retain)	MANSU	Rt54.1	16	19d ago
WMDP-chem (retain)		Rt (Knowledge Retention)56	16	19d ago
WMDP bio (retain)		Rt80.3	16	19d ago
MUSE utility	LUNAR	Util Score18.4	14	19d ago
Complete 30% (retain)	GA_diff	Class QA Accuracy64.64	14	22d ago
Personalized (retain)	MMunlearner	Class VQA Accuracy67.42	14	22d ago
Selective (retain)	GA_diff	Class VQA Accuracy64.39	14	22d ago
Internal e-commerce benchmark Neighbours medium-scale seller 387 items		Rouge Score80.8	14	3mo ago
CIFAR-100 Task-0 (test)	BiCLIP	Accuracy86.2	12	26d ago
Knowledge Retention Suite MME, MMBench, SEEDBench2 Plus, OCRVQA, ScienceQA, MMMU, MIA-Bench, MMDU, MathVista, MathVision, POPE, HallusionBench		COM65.61	12	26d ago
Proposed Driving Benchmark	RecogDrive-8B	KRR74	9	1mo ago
Sequential Unlearning 16-task Retained Data Avg		Specificity100	9	2mo ago
MUSE Retain set (Dr)		KnowMem56	9	3mo ago
Knowledge Retention Evaluation Suite		MME82.54	8	26d ago
MMBench General Knowledge Retention		COM Score81.18	8	26d ago
16-task Sequential Unlearning Retained Data Last	CORE	Specificity96.54	8	2mo ago
internal e-commerce benchmark large-scale seller 1065 items (Neighbours)	ME+GD	Rouge Score58.2	8	3mo ago
RWKU Famous People Neighbor Set	ICU	FB Score56.6	7	3mo ago
Knowledge retention personalization benchmark (test)	Consolidated	Semantic Score94.3	4	8d ago
Llama 8B Pretraining Distribution 3.1	LoRA	Retain (pp)12.3	4	1mo ago

Showing 25 of 25 rows