from Hacker News

nyrikki

joined 8/11/17, 5:40 AM has 1478 karma

• Mind the Gap: Deep Learning Doesn't Learn Deeply

  by nyrikki on 5/30/25, 3:31 PM, with 0 comments

• Beyond Semantics: Unreasonable Effectiveness of Reasonless Intermediate Tokens

  by nyrikki on 5/23/25, 4:13 PM, with 66 comments

• Do Think Tags Help LLMs Plan?

  by nyrikki on 5/14/25, 1:27 PM, with 1 comments

• (How) Do reasoning models reason?

  by nyrikki on 4/21/25, 1:27 PM, with 0 comments