Indie game storeFree gamesFun gamesHorror games
Game developmentAssetsComics
SalesBundles
Jobs
Tags
A jam submission

Iterative summarization interpretabilityView project page

Experiment on Iterative Summarisation, potentially linked to Natural Abstraction. Made for the Alignment jam #4.
Submitted by Yoann Poupart — 3 hours, 52 minutes before the deadline
Add to collection

Play tool

Iterative summarization interpretability's itch.io page

Results

CriteriaRankScore*Raw Score
ML Safety#92.8583.500
Reproducibility#112.8583.500
Generality#132.4493.000
Mechanistic interpretability#132.4493.000
Novelty#132.4493.000

Ranked from 2 ratings. Score is adjusted from raw score by the median number of ratings per game in the jam.

Judge feedback

Judge feedback is anonymous.

  • It's a very interesting piece of work with the use of these two complementary networks as an approach to exploring the Telephone Theorem! A few constructive points: The methods section could be more detailed and specific about the training and evaluation of the networks and the datasets used. The results section could benefit from more quantitative analysis and evaluation of the performance of the networks, with potential for expansion into comparing with human behaviour. But there's definitely potential in continuing this direction, good work.

What are the full names of your participants?
Yoann Poupart

What is you and your team's career stage?
Field discovering

Does anyone from your team want to work towards publishing this work later?

Maybe

Where are you participating from?

Online

Leave a comment

Log in with itch.io to leave a comment.

Comments

No one has posted a comment yet