AI Paper Atlas

Research map / Large language models / Language, safety and interpretability

Leveraging the Inductive Bias of Large Language Models for Abstract Textual Reasoning

Christopher Michael Rytting, David Wingate

NeurIPS 2021 · Large language models › Language, safety and interpretability · approach: Text generation and evaluation metrics · 5 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
Train on the full, nonspecific noun set rather than a 2k common-noun subsetT5-3B container reasoner · Containers, validation words (all)exact: 0.60 -> 0.78Improved
Train on the full, nonspecific noun set rather than a 2k concrete-noun subsetT5-3B container reasoner · Containers, validation words (all)exact: 0.59 -> 0.78Improved
Train on the full, nonspecific noun set rather than 2k randomly sampled nounsT5-3B container reasoner · Containers, validation words (all)exact: 0.55 -> 0.78Improved
Use a larger, nonspecific noun training distributionT5-3B container reasoner · Containers, new POS (random strings)exact: 0.37 -> 0.44; 0.31 -> 0.44; 0.40 -> 0.44Mixed
Train first on elemental navigation and container tasks before the composite taskT5-3B hard object reasoner · Hard object composite taskperformance and sample efficiencyImproved
Train on a path-length-balanced navigation distribution instead of randomly sampling maps and endpointsT5-3B navigation reasoner · Navigation route, four- and five-step plansexact string accuracyImproved
Use a path-length-balanced navigation distributionT5-3B navigation reasoner · Navigation route, two-step plansexact string accuracyHurt
Replace natural English template words with one-to-one gibberish equivalents (ablation)T5-3B container reasoner · Containers, gibberish-template comparisonlearning performanceHurt
Increase the number of objects and containers beyond the training rangeT5-3B container reasoner · Containers, cardinality extrapolationexact string accuracyMixed
Require longer route outputs than seen during trainingT5-3B navigation route reasoner · Navigation route, longer plansexact string accuracyHurt

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy