How Random is Random? Evaluating the Randomness and Humaness of LLMs' Coin Flips
Fuente:
arXiv
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866929369169002496 |
|---|---|
| author | Van Koevering, Katherine Kleinberg, Jon |
| author_facet | Van Koevering, Katherine Kleinberg, Jon |
| contents | One uniquely human trait is our inability to be random. We see and produce patterns where there should not be any and we do so in a predictable way. LLMs are supplied with human data and prone to human biases. In this work, we explore how LLMs approach randomness and where and how they fail through the lens of the well studied phenomena of generating binary random sequences. We find that GPT 4 and Llama 3 exhibit and exacerbate nearly every human bias we test in this context, but GPT 3.5 exhibits more random behavior. This dichotomy of randomness or humaness is proposed as a fundamental question of LLMs and that either behavior may be useful in different circumstances. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2406_00092 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | How Random is Random? Evaluating the Randomness and Humaness of LLMs' Coin Flips Van Koevering, Katherine Kleinberg, Jon Artificial Intelligence Machine Learning One uniquely human trait is our inability to be random. We see and produce patterns where there should not be any and we do so in a predictable way. LLMs are supplied with human data and prone to human biases. In this work, we explore how LLMs approach randomness and where and how they fail through the lens of the well studied phenomena of generating binary random sequences. We find that GPT 4 and Llama 3 exhibit and exacerbate nearly every human bias we test in this context, but GPT 3.5 exhibits more random behavior. This dichotomy of randomness or humaness is proposed as a fundamental question of LLMs and that either behavior may be useful in different circumstances. |
| title | How Random is Random? Evaluating the Randomness and Humaness of LLMs' Coin Flips |
| topic | Artificial Intelligence Machine Learning |
| url | https://arxiv.org/abs/2406.00092 |