AI Paper Atlas

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

Capturing Failures of Large Language Models via Human Cognitive Biases

Erik Jones, Jacob Steinhardt

NeurIPS 2022 · Large language models › Language, safety and interpretability · approach: Statistical frameworks for LLM evaluation · 34 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
prepend irrelevant preceding function (framing line) to promptCodex · HumanEvalfunctional accuracy: -22.3 to -30.5 pointsHurt
prepend irrelevant preceding functionCodex · HumanEvalframing line output rate: 4.5% -> 81%Hurt
prepend irrelevant preceding functionCodeGen · HumanEvalframing line output rate: 0.0% -> 70.7%Hurt
flip operation order from unary-first to binary-firstCodex · MathEquationsaccuracy: 50% -> 17%Hurt
request a contradictory function nameCodex · MathEquationsaccuracy: 100% -> 4.4%Hurt
add anchoring lines from canonical solution to promptCodex · HumanEvalanchor-based generation: for var in 32%-61% of solutionsHurt
anchor numerical estimate prompts on GPT-3GPT-3 · anchoring adaptationestimate shift: no change 10.7 vs 14.3 for 20% vs 50% anchorsMixed

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