SOTA Claims

Extracted SOTA claims and unrestricted claim audit

SOTA Claims

State-of-the-art performance claims extracted from pre-appendix paper prose (not tables or appendix). Section B audits unrestricted claims against at-publication leaderboard data.

Total claims: 10 Unrestricted: 1 Restricted: 7

All extracted claims

Paper	Year	Raw quote	Summary	Datasets	Scope	Restrictions	Source	Confidence
Heterophily-Agnostic Hypergraph Neural Networks with Riemannian Local Exchanger	2026	Based on the proposed mechanism, we propose a novel HealHGNN, a node-hyperedge bidirectional system, which achieves heterophily-agnostic message passing with…	HealHGNN achieves heterophily-agnostic message passing with linear complexity and superior performance on…	—	unrestricted	—	Introduction	high
Heterophily-Agnostic Hypergraph Neural Networks with Riemannian Local Exchanger	2026	Our model achieves competitive performance across all real-world datasets—matching strong baselines on homophilic hypergraphs and delivering clear gains on heterophilic…	HealHGNN matches strong baselines on homophilic datasets and outperforms them on heterophilic…	—	restricted	baseline_subset: performance is compared specifically against homophilic and heterophilic subsets of datasets	Hypergraph Node Classification (RQ1)	high
Heterophily-Agnostic Hypergraph Neural Networks with Riemannian Local Exchanger	2026	Notably, on the Senate dataset we outperform the second-best method by 5%.	HealHGNN outperforms the second-best method on the Senate dataset by 5%.	Senate	restricted	baseline_subset: compared to the second-best method	Hypergraph Node Classification (RQ1)	high
Heterophily-Agnostic Hypergraph Neural Networks with Riemannian Local Exchanger	2026	However, as the heterophily level increases, our approach demonstrates a significant improvement.	HealHGNN shows significant improvement over baselines as heterophily increases in synthetic hypergraphs.	Synthetic Hypergraph	restricted	data_source: synthetic hypergraphs with varying heterophily levels	Synthetic Hypergraph under Heterophily	high
Heterophily-Agnostic Hypergraph Neural Networks with Riemannian Local Exchanger	2026	In contrast, HealHGNN demonstrates consistently strong performance.	HealHGNN shows consistently strong performance on hypergraph transfer tasks.	—	restricted	data_source: hypergraph transfer tasks (HES, HEP, HER, HED)	Hypergraph Transfer (RQ2)	medium
Heterophily-Agnostic Hypergraph Neural Networks with Riemannian Local Exchanger	2026	More importantly, on HEP, HER, and HED, where the structures are more complex and the required propagation distance is longer…	HealHGNN outperforms all other models on complex hypergraph transfer tasks (HEP, HER,…	—	restricted	baseline_subset: compared to all other models on specific complex transfer tasks	Hypergraph Transfer (RQ2)	high
Heterophily-Agnostic Hypergraph Neural Networks with Riemannian Local Exchanger	2026	In contrast, HealHGNN achieves superior performance while maintaining linear complexity (	V	+	E	).	HealHGNN achieves superior performance on the Long-Range Graph Benchmark (LRGB) with linear…	Peptides-func, Peptides-struct
Heterophily-Agnostic Hypergraph Neural Networks with Riemannian Local Exchanger	2026	In contrast, HealHGNN achieves superior performance while maintaining linear complexity (		+		).	HealHGNN achieves superior performance with linear complexity.	—
Heterophily-Agnostic Hypergraph Neural Networks with Riemannian Local Exchanger	2026	As discussed previously, the effect of oversmoothing becomes more severe in heterophilic hypergraphs, amplifying the performance gap in favor of…	HealHGNN shows a larger performance advantage in heterophilic hypergraphs due to better…	House, Senate	restricted	data_source: heterophilic hypergraphs	Neural Network Depth (RQ3)	high
Heterophily-Agnostic Hypergraph Neural Networks with Riemannian Local Exchanger	2026	Extensive experiments across multiple benchmarks demonstrate that HealHGNN consistently improves performance over competitive baselines.	HealHGNN consistently outperforms competitive baselines across multiple benchmarks.	—	unclear	—	Conclusion	high

Unrestricted claim audit

Claims with scope=unrestricted and no listed restrictions, compared to the best known result on each dataset at the paper’s publication date.

Paper	Claim	Dataset	Paper result	Best at pub	Prior best paper	Gap	Verdict
Heterophily-Agnostic Hypergraph Neural Networks with Riemannian Local Exchanger	Based on the proposed mechanism, we propose a novel HealHGNN, a node-hyperedge bidirectional system, which achieves…	—	—	—	—	—	unverifiable (no datasets named in claim)