Our services
Consultancy on LLM evaluation
We provide consultation services on state of the art philosophical foundations (semantic, pragmatic, syntactic, logical and epistemological) for the evaluation, benchmarking and ranking of major language models.
Training in LLM benchmarking fundamentals
We offer training in the conceptual underpinnings of LLM benchmarking. We help teams grasp the current conceptual frameworks used to benchmark LLMs, which involves understanding the major types of benchmarks, their uses, and the current trends and limitations involving them.This service is meant for LLM research and development teams, AI engineering teams, and actors in the corporate or public sector interested in implementing their own benchmarks or learning how to read them.
Tailored solutions on LLM benchmarking
We offer guidance on the implementation of in-house or public benchmarks that require adversarial, synthetic or expert tailored data in the areas of language understanding, reasoning, theory of mind, and related phenomena.