M

Group

What will be the state-of-the-art language modelling performance (in perplexity) on WikiText-103 by the following dates?

30
103 forecasters

Make a Prediction

Dateresolution
Date25thmedian75th
community
<8
lower 25%
median
upper 75%
0.5%
14.59
15.33
15.69

What was the final result?14.8

Community Baseline Score
35.0
Community Peer Score
32.5
Learn more about scores here.

Forecast Timeline

Total Forecasters 44
Authors:
casens
Opened:
Closes:
Scheduled resolution:
Forecasting AI Progress
Forecasting AI Progress: Maximum Likelihood Round
AI Progress Essay Contest
AI Technical Benchmarks
Forecasting AI Progress: Hill Climbing Round
Computing and Math
Artificial Intelligence
2021-2025 Leaderboard
2021 Leaderboard
2020-2021 Leaderboard
2016-2025 Leaderboard
Machine learning
Linguistics
Information technology

Comments

? comments