CSPaper: sidekick of peer review

KDD reviews are coming out, and most community-shared scores cluster around 2.6 to 3.6 on average. Common patterns are like 333xx, 433xx, or 422xx, showing that many papers are seen as average in novelty and technical quality. Even submissions authors are proud of are mostly getting 3s, with only a few 4s or 5s. The overall vibe: “low scores are normal, let’s just hope for kind reviewers and make the most of rebuttal.”

In short, KDD remains tough, and scores are modest across the board.

Even if a rebuttal is detailed and thorough, reviewers often only ACK without changing the score. This usually means they accept your response but don’t feel it shifts their overall assessment enough. Some see added experiments as “too late” or not part of the original contribution. Others may still not fully understand the paper but won’t admit it. Unfortunately, rebuttals prevent score drops more often than they raise scores.

My initial scores are OA 2.5/2.5/2
It got raised to OA 2.5/2.5/2.5 ... Well, become a bit better LOL!
Any chance for findings?

Table 1: Comparison to the most recent baselines available at https://github.com/thuml/Time-Series-Library