Ori Plonsky

Asst Prof @TechnionLive.
Behavioral science | Data science | Behavioral economics | Predicting behavior | Decision making.

[Ori Plonsky @OriPlonsky]

Proud & excited to share, 8 years(!) after starting this work.

🚨New in @NatureHumBehav🚨
BEAST-GB: a hybrid model predicting human choice at near-ceiling levels.

It uses small data, beats behavioral & AI models, generalizes across contexts, and even explains human choice!

1/x

[Nature Human Behaviour @NatureHumBehav]

In this Article, @OriPlonsky et al. introduce a new model that merges behavioral science and machine learning to predict choice under risk and uncertainty. Tested on multiple large datasets, it exceeds top psychological and AI models.
https://t.co/1WDoYQBY8t

[Ori Plonsky @OriPlonsky]

@the_yaniv ציר x ארביטררי (לא ארביטררי)

[Ori Plonsky @OriPlonsky]

Shout-out to my awesome co-authors - @reut_apel, Eyal Ert, @moshetenn, David Bourgin, Josh C Peterson, Daniel Reichman, @cocosci_lab, Stuart J Russel, Even C Carter, James F Cavanagh, and Ido Erev.

10/9

[Ori Plonsky @OriPlonsky]

Proud & excited to share, 8 years(!) after starting this work.

🚨New in @NatureHumBehav🚨
BEAST-GB: a hybrid model predicting human choice at near-ceiling levels.

It uses small data, beats behavioral & AI models, generalizes across contexts, and even explains human choice!

1/x

[Nature Human Behaviour @NatureHumBehav]

In this Article, @OriPlonsky et al. introduce a new model that merges behavioral science and machine learning to predict choice under risk and uncertainty. Tested on multiple large datasets, it exceeds top psychological and AI models.
https://t.co/1WDoYQBY8t

[Ori Plonsky @OriPlonsky]

Much more analysis in the paper, e.g., similar hybrid models predict risky individual choice and even strategic choice.

We believe the same approach may help in much more complex domains (stay tuned).

Thx for reading and sharing!

Full paper: https://t.co/KfPNeW4syf
9/9

[Eugen Dimant @eugen_dimant]

🔥New publication on dishonesty🔥

In a new collaborative paper (led by Shaul Shalvi), we review the key interdisciplinary frameworks on (dis)honesty & offer 66 open questions to guide future work

🔗(OA) https://t.co/63NkGSwoKp

@IsabelThielmann @ValerioCapraro @JF_Schulz et al.

[Ori Plonsky @OriPlonsky]

✅The results of the competition are proof of concept that proper quantitative behavior models can be used not only to describe or predict, but also to engineer behavior.

[Ori Plonsky @OriPlonsky]

🧠 Choice architecture + 📊 Computational models = 🛠️ **Choice Engineering**

Proud to share our new paper published in Nature Communications
@NatureComms
📄 "Behavior engineering using quantitative reinforcement learning models".

https://t.co/8scBHbykQx
[1/n]

[Ori Plonsky @OriPlonsky]

🎯CATIE convinced participants to choose the targeted option in ~65% of decisions, despite identical outcomes!

Unlike classic RL, CATIE assumes people try to follow sequential patterns rather than updating option-values based on past outcomes. (https://t.co/VL3bRf3drC…)
[5/n]

[Ori Plonsky @OriPlonsky]

@liadlitman @DeanBracha אוקיי, בדקתי שוב ואתה צודק.
(בהרחבה: הצהובים באמת נמחקים בסוף רבעי הגמר, אבל אם שחקן מקבל צהוב שלישי לפני סוף רבע הגמר, אז הוא מושעה מהמשחק הבא לפני שהצהובים שלו נמחקים...)

[Ori Plonsky @OriPlonsky]

@liadlitman @DeanBracha לא. צהובים נמחקים בסוף המשחק