AI Paper Atlas

Research map / Large language models / Agents, code and math

Mind the Gap: The Divergence Between Human and LLM-Generated Tasks

Yi-Long Lu, Jiajun Song, Chunhui Zhang, Wei Wang

AAAI 2026 · Large language models › Agents, code and math · approach: Agent instruction-following benchmarks · 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
condition GPT-4o on matched human psychological profile (values, cognitive style) (ablation)GPT-4o task generation · Task theme distributionshare of Relaxation & Household tasks: 3% (raw) vs 5% (matched) vs 32% (humans)No clear effect
condition GPT-4o on matched human profile (ablation)GPT-4o task generation · Multiplayer tasks (another person present)share of multiplayer tasks: 29% -> 24% (humans 58%)No clear effect
higher Openness to Change value in human participantshuman task generation · Human task ratingstask novelty rating: b = 0.152, p = 0.017Improved
higher Openness to Change value in human participantshuman task generation · Human task ratingstask fun rating: b = 0.146, p = 0.022Improved
self-interest value dimension as predictorhuman task generation · Human task ratingsnovelty and funNo clear effect
more systematic thinking style (TWS x environmental complexity interaction)human task generation · Human task ratingstask fun: interaction b = -0.140, p = 0.009Mixed
more systematic thinking style (TWS x environmental complexity interaction)human task generation · Human task diversitytask diversity (cosine distance): interaction b = -0.215, p = 0.006Mixed
use GPT-4o (matched) instead of humanstask generation · Rated mental demandmental demand rating: z = -11.64, pFDR < 0.001Hurt
use GPT-4o (matched) instead of humanstask generation · Rated physical demandphysical demand rating: z = -8.40, pFDR < 0.001Hurt
use GPT-4o (matched) instead of humanstask generation · Rated noveltynovelty rating: z = -5.63, pFDR < 0.001Improved
use GPT-4o (matched) instead of humanstask generation · Rated funfun rating: z = -2.73, pFDR = 0.010Improved
use GPT-4o (matched) instead of humanstask generation · Body part engagementarms and lower body engagement: arms z = -5.43; lower body z = -7.93Hurt

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