Methods › General › Negative Sampling › Self-Adversarial Negative Sampling
Self-Adversarial Negative Sampling
archive 2025-07-28 Description, source and code snippet are the archive's method entry.
Self-Adversarial Negative Sampling is a negative sampling technique used for methods like word embeddings and knowledge graph embeddings. The traditional negative sampling loss from word2vec for optimizing distance-based models be written as:
L = −logσ(γ− dᵣ(𝐡, 𝐭)) − ∑ⁿᵢ₌₁1/klogσ(dᵣ(𝐡^′ᵢ, 𝐭^′ᵢ) - γ)
where γ is a fixed margin, σ is the sigmoid function, and (𝐡^′ᵢ, r, 𝐭^′ᵢ) is the i-th negative triplet.
The negative sampling loss samples the negative triplets in a uniform way. Such a uniform negative sampling suffers the problem of inefficiency since many samples are obviously false as training goes on, which does not provide any meaningful information. Therefore, the authors propose an approach called self-adversarial negative sampling, which samples negative triples according to the current embedding model. Specifically, we sample negative triples from the following distribution:
p(h^′ⱼ, r, t^′ⱼ | set(hᵢ, rᵢ, tᵢ ) ) = (expαfᵣ(𝐡^′ⱼ, 𝐭^′ⱼ))/(∑ᵢ₌₁expαfᵣ(𝐡^′ᵢ, 𝐭^′ᵢ))
where α is the temperature of sampling. Moreover, since the sampling procedure may be costly, the authors treat the above probability as the weight of the negative sample. Therefore, the final negative sampling loss with self-adversarial training takes the following form:
L = −logσ(γ− dᵣ(𝐡, 𝐭)) − ∑ⁿᵢ₌₁p(h^′ᵢ, r, t^′ᵢ)logσ(dᵣ(𝐡^′ᵢ, 𝐭^′ᵢ) - γ)
Papers archive 2025-07-28
30 shown of 30, newest first. Repository counts are the archive's code-links table. A Syntology line states what Syntology ran from that paper's harvested code; it is per sample and not a correctness claim.
-
Is Architectural Complexity Overrated? Competitive and Interpretable Knowledge Graph Completion with RelatE 25 May 2025 · 0 repositories · arXiv:2505.18971
-
SparseTransX: Efficient Training of Translation-Based Knowledge Graph Embeddings Using Sparse Matrix Operations 24 Feb 2025 · 2 repositories · arXiv:2502.16949
-
Unified Interpretation of Smoothing Methods for Negative Sampling Loss Functions in Knowledge Graph Embedding 5 Jul 2024 · 1 repository · arXiv:2407.04251
-
PAC-Bayesian Generalization Bounds for Knowledge Graph Representation Learning 10 May 2024 · 1 repository · arXiv:2405.06418Syntology ran 3 of 4 samples · 1 unverified · 4 pointer-only (licence)
-
Edge-Enabled Anomaly Detection and Information Completion for Social Network Knowledge Graphs 13 Jan 2024 · 0 repositories · arXiv:2401.07022
-
Block-Diagonal Orthogonal Relation and Matrix Entity for Knowledge Graph Embedding 11 Jan 2024 · 1 repository · arXiv:2401.05967
-
Does Pre-trained Language Model Actually Infer Unseen Links in Knowledge Graph Completion? 15 Nov 2023 · 0 repositories · arXiv:2311.09109
-
Model-based Subsampling for Knowledge Graph Completion 17 Sep 2023 · 1 repository · arXiv:2309.09296
-
BESS: Balanced Entity Sampling and Sharing for Large-Scale Knowledge Graph Completion 22 Nov 2022 · 2 repositories · arXiv:2211.12281Syntology ran 0 of 9 samples · 9 unverified
-
Self-Improving Safety Performance of Reinforcement Learning Based Driving with Black-Box Verification Algorithms 29 Oct 2022 · 2 repositories · arXiv:2210.16575
-
CompoundE: Knowledge Graph Embedding with Translation, Rotation and Scaling Compound Operations 12 Jul 2022 · 0 repositories · arXiv:2207.05324
-
Comprehensive Analysis of Negative Sampling in Knowledge Graph Representation Learning 21 Jun 2022 · 1 repository · arXiv:2206.10140
-
Learning to Borrow -- Relation Representation for Without-Mention Entity-Pairs for Knowledge Graph Completion 27 Apr 2022 · 1 repository · arXiv:2204.13097
-
SpaceE: Knowledge Graph Embedding by Relational Linear Transformation in the Entity Space 21 Apr 2022 · 0 repositories · arXiv:2204.10245
-
A Unified Framework for Rank-based Evaluation Metrics for Link Prediction in Knowledge Graphs 14 Mar 2022 · 2 repositories · arXiv:2203.07544Syntology ran 0 of 5 samples · 5 unverified
-
SimKGC: Simple Contrastive Knowledge Graph Completion with Pre-trained Language Models 4 Mar 2022 · 2 repositories · arXiv:2203.02167Syntology ran 3 of 6 samples · 3 unverified · 3 pointer-only (licence)
-
Learning to Borrow– Relation Representation for Without-Mention Entity-Pairs for Knowledge Graph Completion 16 Jan 2022 · 0 repositories
-
CORE: A Knowledge Graph Entity Type Prediction Method via Complex Space Regression and Embedding 19 Dec 2021 · 0 repositories · arXiv:2112.10067
-
Mixture-of-Graphs: Zero-shot Relational Learning for Knowledge Graph by Fusing Ontology and Textual Experts 16 Nov 2021 · 0 repositories
-
QubitE: Qubit Embedding for Knowledge Graph Completion 16 Nov 2021 · 1 repository
-
Towards Robust Knowledge Graph Embedding via Multi-task Reinforcement Learning 11 Nov 2021 · 0 repositories · arXiv:2111.06103
-
Why does Negative Sampling not Work Well? Analysis of Convexity in Negative Sampling 29 Sep 2021 · 0 repositories
-
NOTE: Solution for KDD-CUP 2021 WikiKG90M-LSC 5 Jul 2021 · 0 repositories · arXiv:2107.01892
-
QuatDE: Dynamic Quaternion Embedding for Knowledge Graph Completion 19 May 2021 · 1 repository · arXiv:2105.09002
-
DensE: An Enhanced Non-commutative Representation for Knowledge Graph Embedding with Adaptive Semantic Hierarchy 11 Aug 2020 · 1 repository · arXiv:2008.04548
-
Convolutional Complex Knowledge Graph Embeddings 7 Aug 2020 · 2 repositories · arXiv:2008.03130
-
Orthogonal Relation Transforms with Graph Context Modeling for Knowledge Graph Embedding 9 Nov 2019 · 0 repositories · arXiv:1911.04910
-
TransGCN:Coupling Transformation Assumptions with Graph Convolutional Networks for Link Prediction 1 Oct 2019 · 0 repositories · arXiv:1910.00702
-
Group Representation Theory for Knowledge Graph Embedding 11 Sep 2019 · 0 repositories · arXiv:1909.05100
-
RotatE: Knowledge Graph Embedding by Relational Rotation in Complex Space 26 Feb 2019 · 10 repositories · arXiv:1902.10197Syntology ran 0 of 5 samples · 5 unverified
Tasks archive 2025-07-28
20 shown of 51 tasks the archive attaches to papers tagged with this method, by distinct papers. A task without a page in the catalog is plain text.
Usage over time archive 2025-07-28
Components: the archive holds no method-to-method composition, so PwC's Components table cannot be rebuilt; the Papers list carries no Results column for the same reason (the archive does not join its leaderboard rows to method tags).
Categories archive 2025-07-28
Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections