Questions to research and think

Liang-Jiaying / RLAIF

MIT License

2 stars 0 forks source link

Open Liang-Jiaying opened 10 months ago

Liang-Jiaying commented 10 months ago

[ ] Why the author only compare RLAIF with RLHF on task of summarization?
[ ] How are the performances for other tasks?
[ ] For 4.1 Datasets, what other ways OpenAI use to filter the data?
[ ] For 4.1 Datasets, why "only posts where the summaries contain between 24 and 48 tokens are included"?
[ ] For 4.1 Datasets, does 123169 posts seems too small to train?

Liang-Jiaying commented 10 months ago

[x] Where the summaries in TL;DR coming from? (Might just provided by authors of posts)