AI Paper Atlas

Research map / Bandits and online learning / Multi-armed bandits

Why so gloomy? A Bayesian explanation of human pessimism bias in the multi-armed bandit task

Dalin Guo, Angela J. Yu

NeurIPS 2018 · Bandits and online learning › Multi-armed bandits · approach: Non-stationary and structured-reward bandits · 3 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
use dynamic belief model instead of fixed belief modellearning model · Human bandit choicespredictive accuracy: 0.3856 -> 0.4182Improved
use dynamic belief model instead of RLlearning model · Human bandit choicespredictive accuracy: 0.4038 -> 0.4182Improved
assume environment volatility (gamma<1) rather than stationarityDBM vs FBM · Prior mean recoverydifference of recovered prior mean across environments: p = 0.4350Improved
vary environment abundance and variancehuman participants · Reported reward estimatesaverage reported estimate: p = 0.1570No clear effect
high versus low abundance environmentDBM · Fitted parameterssoftmax b and gamma: p < .01Mixed
high versus low variance environmentDBM · Fitted parameterssoftmax b and gamma: p > .05No clear effect
fit parameters per participant and environment instead of shared (ablation)DBM/FBM/RL fitting · 10-fold cross validationheld-out predictive accuracyImproved
underestimate prior mean in FBM with softmax policyFBM · Simulated high variance environmentsearned reward per trialImproved
underestimate prior mean even more in DBMDBM · Simulated high variance environmentsearned reward per trialImproved

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