#KostasThoughts: If a paper promises code/weights in the review submission, then it must be made available by the time of the conference; otherwise, conferences should desk reject the paper.
11:43 PM · Jun 7, 2026 · 508 Views
Kosta Derpanis, a York University associate professor, proposes academic conferences desk-reject papers that fail to release promised code or model weights
NeurIPS already requires code and weights at submission time.
455312.8K
Sentiment
Sentiment building, check back later.
Cluster Engagement
Posts from X
Most Activity
Most Activity
VIEWS2.3KBOOKMARKS1LIKES50RETWEETS3REPLIES4
Kosta Derpanis (sabbatical in Zurich)@CSProfKGD
#KostasThoughts: If a paper promises code/weights in the review submission, then it must be made available by the time of the conference; otherwise, conference should desk reject the paper.
3hViews 2.3KLikes 50Bookmarks 1
Francesco Locatello@FrancescoLocat8
@CSProfKGD In the last two years of the NeuIPS datasets and benchmarks track, we required it at submission if they were contributions. The rationale was that whatever is a contribution should be reviewed. Same as a theory paper, which we expect to have the proof
Kosta Derpanis (sabbatical in Zurich)@CSProfKGD
#KostasThoughts: If a paper promises code/weights in the review submission, then it must be made available by the time of the conference; otherwise, conference should desk reject the paper.
1hViews 140Likes 3Bookmarks 0
antonio vergari ⚔️@tetraduzione
this
Kosta Derpanis (sabbatical in Zurich)@CSProfKGD
#KostasThoughts: If a paper promises code/weights in the review submission, then it must be made available by the time of the conference; otherwise, conference should desk reject the paper.
35mViews 79Likes 1Bookmarks 0