Reproduce “Predicting Linear Relational Decodability Without the Relation” on OpenResearch

Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.

Universal app for Apple silicon and Intel. Requires macOS 11 or later.

About this paper

Jonas Rohweder, Adrian Mülthaler, Faridoun Mehri, Sebastian Gerstner, Hinrich Schütze

The authors test whether a relation-specific linear relational embedding can reproduce a language model’s answer from a subject representation, and whether a neutral entity lens predicts its faithfulness. The lens reads what the model evokes about same-type entities without the relation prompt. Across 28 OLMo-3-7B relations, scores correlate (Spearman ρ = 0.7355); this predicts agreement with the model, not factual accuracy.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.