Reasoning and Tools for Human-Level Forecasting

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Hsieh, Elvis, Fu, Preston, Chen, Jonathan
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912099351920640
author Hsieh, Elvis
Fu, Preston
Chen, Jonathan
author_facet Hsieh, Elvis
Fu, Preston
Chen, Jonathan
contents Language models (LMs) trained on web-scale datasets are largely successful due to their ability to memorize large amounts of training data, even if only present in a few examples. These capabilities are often desirable in evaluation on tasks such as question answering but raise questions about whether these models can exhibit genuine reasoning or succeed only at mimicking patterns from the training data. This distinction is particularly salient in forecasting tasks, where the answer is not present in the training data, and the model must reason to make logical deductions. We present Reasoning and Tools for Forecasting (RTF), a framework of reasoning-and-acting (ReAct) agents that can dynamically retrieve updated information and run numerical simulation with equipped tools. We evaluate our model with questions from competitive forecasting platforms and demonstrate that our method is competitive with and can outperform human predictions. This suggests that LMs, with the right tools, can indeed think and adapt like humans, offering valuable insights for real-world decision-making.
format Preprint
id arxiv_https___arxiv_org_abs_2408_12036
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Reasoning and Tools for Human-Level Forecasting
Hsieh, Elvis
Fu, Preston
Chen, Jonathan
Machine Learning
Artificial Intelligence
Computation and Language
Information Retrieval
Language models (LMs) trained on web-scale datasets are largely successful due to their ability to memorize large amounts of training data, even if only present in a few examples. These capabilities are often desirable in evaluation on tasks such as question answering but raise questions about whether these models can exhibit genuine reasoning or succeed only at mimicking patterns from the training data. This distinction is particularly salient in forecasting tasks, where the answer is not present in the training data, and the model must reason to make logical deductions. We present Reasoning and Tools for Forecasting (RTF), a framework of reasoning-and-acting (ReAct) agents that can dynamically retrieve updated information and run numerical simulation with equipped tools. We evaluate our model with questions from competitive forecasting platforms and demonstrate that our method is competitive with and can outperform human predictions. This suggests that LMs, with the right tools, can indeed think and adapt like humans, offering valuable insights for real-world decision-making.
title Reasoning and Tools for Human-Level Forecasting
topic Machine Learning
Artificial Intelligence
Computation and Language
Information Retrieval
url https://arxiv.org/abs/2408.12036