We propose Semantic Averaging with MUltiple Languages (SAMUL) as a method to isolate concept semantics of text by averaging concept activations derived via Sparse Autoencoders (SAEs) for multiple text versions in different languages. We combine autoencoder activations to derive a conceptual average using different approaches. In our experiments with SAMUL, we start with English text from benchmark datasets (Ontology Alignment Evaluation Initiative 2025: conference track, SemEval2014: cross-level semantic similarity, Sentences Involving Compositional Knowledge: relatedness) and automatically translate into French, Chinese, German and Japanese. We obtain SAE feature activations for each class and language version using autoencoders, specifically the open source Gemma Scope 1 and OpenAI’s GPT2 suites of Sparse Autoencoders and average them to SAMUL vectors. We measure performance in predicting semantic similarity of short texts using predefined ground truth or correlation to the ground truth mapping m between ontology classes in the case of the ontology alignment task. Our results show a consistent improvement over any single translated language, and improvement over English alone in 27 of 32 comparisons. Conceptual averaging therefore yields sparse representations that align more closely with human similarity judgements, suggesting a route to more reliable use of SAE features in semantic tasks such as ontology alignment. Our implementation and experimentation code is openly available.

There are many methods and systems to tackle the ontology alignment problem, yet a major challenge persists in producing high-quality mappings among a set of input ontologies. Adopting a human-in-the-loop approach during the alignment process has become essential in applications requiring very accurate mappings. However, user involvement is expensive when dealing with large ontologies. In this paper, we analyse the feasibility of using Large Language Models (LLM) to aid the ontology alignment problem. LLMs are used only in the validation of a subset of correspondences for which there is high uncertainty. We have conducted an extensive analysis over several tasks of the Ontology Alignment Evaluation Initiative (OAEI), reporting in this paper the performance of several state-of-the-art LLMs using different prompt templates. Using LLMs as resulted in strong performance in the OAEI 2025, achieving the top-2 overall rank in the bio-ml track.

Investigating whether pre-trained language models (LMs) can function as knowledge bases (KBs) has raised wide research interests recently. However, existing works focus on simple, triple-based, relational KBs, but omit more sophisticated, logic-based, conceptualised KBs such as OWL ontologies. To investigate an LM’s knowledge of ontologies, we propose OntoLAMA, a set of inference-based probing tasks and datasets from ontology subsumption axioms involving both atomic and complex concepts. We conduct extensive experiments on ontologies of different domains and scales, and our results demonstrate that LMs encode relatively less background knowledge of Subsumption Inference (SI) than traditional Natural Language Inference (NLI) but can improve on SI significantly when a small number of samples are given. We will open-source our code and datasets.