Hey all, We have received way too many submissions for our workshop at NeurIPS and we are looking for reviewers. If you can help review a paper, within next month, then please let us know your Openreview ID!

Show Post

Joint work with Ayan Majumdar, Shounak Paul, Pushpdeep Singh, Ines Abdelaziz, Sayeh Jarollahi, Seungeon Lee, Krishna P. Gummadi, Ingmar Weber.

Show Post

Our results suggest that they are a promising building block, rather than a replacement for human moderators. But important challenges remain (e.g., biases, scalability etc.) to be explored in deeper details.

Show Post

Another alternative is to provide the VLMs with prior moderation decisions as examples. The best example-driven approach achieves nearly identical F1 performance. This parity comes at a severe practical cost in latency and expenditure.

Show Post

Within the instruction-driven paradigm, granu- lar policy details are critical in boosting moderation consistency and effectiveness: roughly halving flag- ging rates and improving F1 scores by up to 25%, indicating the importance of better harness or policy for better operationalization.

Show Post