AI Paper Atlas

Research map / Benchmarks / FB15k-237

FB15k-237 leaderboard

58 papers at ICML, NeurIPS, ICLR, CVPR, AAAI, ICCV and ECCV, and the journals TPAMI, IJCV and JMLR report results on FB15k-237. Best MRR: 0.53 (Ours 3xP100+2xD100, NeurIPS 2023).

Numbers are as each paper reports them, read by AI from its results tables and not checked by hand. Papers can differ in setting (training data, model size, extra data); the topic column shows what each paper is about. Open the paper before you compare or quote a number.

FB15k-237 · MRR

55 papers, each with its best reported number for its own method; higher is better.

#MRR ↑MethodPaperTopic
10.53Ours 3xP100+2xD100
RotX
Data-Informed Geometric Space Selection
NeurIPS 2023 · Table 1
Molecules and proteins
20.522RGHATRelational Graph Neural Network with Hierarchical Attention for Knowledge Graph Completion
AAAI 2020 · Table 4
Knowledge and temporal graphs
30.52KGDMKGDM: A Diffusion Model to Capture Multiple Relation Semantics for Knowledge Graph Embedding
AAAI 2024 · Table 2
Knowledge and temporal graphs
40.453DuetGraphDuetGraph: Coarse-to-Fine Knowledge Graph Reasoning with Dual-Pathway Global-Local Fusion
NeurIPS 2025 · Transductive performance table
Knowledge and temporal graphs
50.43KnowFormerKnowFormer: Revisiting Transformers for Knowledge Graph Reasoning
ICML 2024 · Transductive results table
Knowledge and temporal graphs
60.428KGCRR (CRR + CompoundE)
CompoundE
KGCRR: An Effective Metric-Driven Knowledge Graph Completion Framework by Designing a Novel Upper Bound Function with Adaptive Approximation to Reciprocal Rank
AAAI 2025 · Table 2
Knowledge and temporal graphs
70.423GaussianPathGaussianPath:A Bayesian Multi-Hop Reasoning Framework for Knowledge Graph Reasoning
AAAI 2021 · Table 3
Visual and representation RL
80.415MKGL
Llama-2-7b-chat
MKGL: Mastery of a Three-Word Language
NeurIPS 2024 · KG completion results table
Retrieval and knowledge
90.415NBFNet (6 layers)Neural Bellman-Ford Networks: A General Graph Neural Network Framework for Link Prediction
NeurIPS 2021 · Ablation study, number of laye
Knowledge and temporal graphs
100.404DANS-KGCDANS-KGC: Diffusion Based Adaptive Negative Sampling for Knowledge Graph Completion
AAAI 2026 · Table 2
Knowledge and temporal graphs
45 more papers on this leaderboard, the baselines reported on it and a CSV download are free with a scime account.
Sign in free to see all

FB15k-237 · Hits@10

10 papers, each with its best reported number for its own method; higher is better.

#Hits@10 ↑MethodPaperTopic
10.576ParamE-CNNParamE: Regarding Neural Network Parameters as Relation Embeddings for Knowledge Graph Completion
AAAI 2020 · Table 3
Knowledge and temporal graphs
20.572KG-FIT-HAKE (LHR)
HAKE
KG-FIT: Knowledge Graph Fine-Tuning Upon Open-World Knowledge
NeurIPS 2024 · Link Prediction Performance Co
Retrieval and knowledge
30.562SPACSPAC: Sparse Partitioning and Adaptive Core Tensor Pruning Model for Knowledge Graph Completion
AAAI 2025 · Table 1
Knowledge and temporal graphs
40.551Synaptic delay (GRSNN)
SNN
Temporal Spiking Neural Networks with Synaptic Delay for Graph Reasoning
ICML 2024 · Table: Results of knowledge gr
Knowledge and temporal graphs
50.549SE-GNNHow Does Knowledge Graph Embedding Extrapolate to Unseen Data: A Semantic Evidence View
AAAI 2022 · Table 1
Knowledge and temporal graphs
60.54GGPNMulti-Relational Graph Representation Learning with Bayesian Gaussian Process Network
AAAI 2022 · Table 1
Knowledge and temporal graphs
70.54SACNEnd-to-End Structure-Aware Convolutional Networks for Knowledge Base Completion
AAAI 2019 · Link prediction table
Knowledge and temporal graphs
80.538BoxE(a)BoxE: A Box Embedding Model for Knowledge Base Completion
NeurIPS 2020 · Table 1
Molecules and proteins
90.535InteractEInteractE: Improving Convolution-Based Knowledge Graph Embeddings by Increasing Feature Interactions
AAAI 2020 · Table 1
Knowledge and temporal graphs
100.491Full ConvEConvolutional 2D Knowledge Graph Embeddings
AAAI 2018 · Table: Ablation study for FB15
Knowledge and temporal graphs

Not affiliated with any conference, journal or the benchmark's authors. Corrections: contact@latenticlabs.com.