Cross-Model Humor Preference Modeling with Cards Against Humanity
Quick summary
arXiv:2608.07481v1 Announce Type: cross Abstract: This paper investigates whether one large language model can approximate the humor preferences of another in a controlled Cards Against Humanity-style task. Two models - GPT-4o as Czar and Claude Opus-4.5 as Player - are evaluated on a binary humor-selection task constructed so that success cannot follow from self-preference. A reflected-cell stability procedure isolates 244 hands on which the two models hold deterministic but opposite preferences, partitioned into a 97-hand context pool and a 147-hand held-out test pool. The Player is then eva
Key takeaways
- arXiv:2608.07481v1 Announce Type: cross Abstract: This paper investigates whether one large language model can approximate the humor preferences of another in a controlled Cards Against Humanity-style task.
- Two models - GPT-4o as Czar and Claude Opus-4.5 as Player - are evaluated on a binary humor-selection task constructed so that success cannot follow from self-preference.
- A reflected-cell stability procedure isolates 244 hands on which the two models hold deterministic but opposite preferences, partitioned into a 97-hand context pool and a 147-hand held-out test pool.
Why it matters
This model development creates a new option for users and a new testing obligation for developers. A fixed evaluation set comparing quality, cost and failure behavior is more useful than launch claims.
Member comments