Summary of Adaptive Sampling Policies Imply Biased Beliefs: a Generalization Of the Hot Stove Effect, by Jerker Denrell
Adaptive Sampling Policies Imply Biased Beliefs: A Generalization of the Hot Stove Effect
by Jerker Denrell
First submitted to arxiv on: 3 Apr 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Machine Learning (stat.ML)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The paper explores the “Hot Stove Effect,” a phenomenon where learning algorithms tend to overcorrect errors by avoiding alternatives with negative estimates, but undercorrect errors by failing to adjust for overestimation. The authors generalize this effect to settings where learners select fewer alternatives based on negative estimates, but do not entirely avoid them. They demonstrate that the negativity bias remains in this setup and also show that Bayesian learners tend to underestimate expected values. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The Hot Stove Effect is a new way of understanding how we learn. It says that when we’re trying to figure something out, we often make mistakes by being too optimistic or pessimistic. This paper shows that this bias can happen even when we’re not avoiding things we don’t like, but still choosing fewer options because they seem less good. The authors prove that this effect happens and also show that some types of learners are more likely to be overly negative. |
