Free · Tech & Psychology

You are not bad at reading research papers.

They are written to survive peer review, not to be read by you. So pick a paper, tell it who you are, and read it at your level. Every paper gets an honest does it hold up?

Tech

Attention Is All You Need

Ashish Vaswani, Noam Shazeer, Niki Parmar, et al. · 2017

Drop the step-by-step reading of older models and let every word attend to every other word at once. That one change, the Transformer, became the foundation of modern language models.

Read it your way →
Tech

Constitutional AI: Harmlessness from AI Feedback

Yuntao Bai, Saurav Kadavath, Sandipan Kundu, et al. · 2022

Teach an AI to be harmless by giving it a short written set of principles (a constitution) and having it critique and revise its own answers, instead of relying on humans to label every harmful response.

Read it your way →
Psychology

Estimating the Reproducibility of Psychological Science

Open Science Collaboration (Brian Nosek et al.) · 2015

A huge team re-ran 100 published psychology studies. Fewer than half reproduced, and the effects that did were about half the original size. This is the paper that made the replication crisis impossible to ignore.

Read it your way →
Tech

Language Models are Few-Shot Learners (GPT-3)

Tom B. Brown, Benjamin Mann, Nick Ryder, et al. · 2020

Make a language model enormous (175 billion parameters) and it starts doing new tasks from just a few examples in the prompt, no retraining. This paper is why prompting became the way we use AI.

Read it your way →
Psychology

Power Posing: Brief Nonverbal Displays Affect Neuroendocrine Levels and Risk Tolerance

Dana R. Carney, Amy J.C. Cuddy, Andy J. Yap · 2010

The famous claim that standing in a powerful pose for two minutes changes your hormones and makes you bolder. It became a sensation, then largely failed to replicate. A useful case study in reading a paper critically.

Read it your way →
Psychology

Revisiting the Marshmallow Test: A Conceptual Replication

Tyler W. Watts, Greg J. Duncan, Haonan Quan · 2018

The famous finding that a child's ability to delay eating a marshmallow predicts lifelong success mostly melts away once you account for the child's background. Willpower mattered far less than circumstances.

Read it your way →