Multi-Agent Comedy Club: Investigating Community Discussion Effects on LLM Humor Generation

Research Summary

Contribution Summary

• Prior work has explored multi-turn interaction and feedback for LLM writing, but evaluations still largely center on prompts and localized feedback, leaving persistent public reception in online communities underexamined.
• We test whether broadcast community discussion improves stand-up comedy writing in a controlled multi-agent sandbox: in the discussion condition, critic and audience threads are recorded, filtered, stored as social memory, and later retriev
• Across 50 rounds (250 paired monologues) judged by five expert annotators using A/B preference and a 15-item rubric, discussion wins 75.6% of instances and improves Craft/Clarity (Δ = 0.440) and Social Response (Δ = 0.422), with occasional

Why It Matters For Eval

• Prior work has explored multi-turn interaction and feedback for LLM writing, but evaluations still largely center on prompts and localized feedback, leaving persistent public reception in online communities underexamined.
• We test whether broadcast community discussion improves stand-up comedy writing in a controlled multi-agent sandbox: in the discussion condition, critic and audience threads are recorded, filtered, stored as social memory, and later retriev

Related Papers

• DesignBench: A Comprehensive Benchmark for MLLM-based Front-end Code Generation Embedding Neighbor
• VQEL: Enabling Self-Play in Emergent Language Games via Agent-Internal Vector Quantization Embedding Neighbor
• RPTS: Tree-Structured Reasoning Process Scoring for Faithful Multimodal Evaluation Embedding Neighbor
• How Well Can LLM Agents Simulate End-User Security and Privacy Attitudes and Behaviors? Embedding Neighbor
• Large Language Models Persuade Without Planning Theory of Mind Embedding Neighbor
• Thinking by Subtraction: Confidence-Driven Contrastive Decoding for LLM Reasoning Embedding Neighbor
• ErrorMap and ErrorAtlas: Charting the Failure Landscape of Large Language Models Embedding Neighbor
• Anatomy of Capability Emergence: Scale-Invariant Representation Collapse and Top-Down Reorganization in Neural Networks Embedding Neighbor
• ContextRL: Enhancing MLLM's Knowledge Discovery Efficiency with Context-Augmented RL Embedding Neighbor
• HEART: A Unified Benchmark for Assessing Humans and LLMs in Emotional Support Dialogue Tag Overlap
• The Subjectivity of Respect in Police Traffic Stops: Modeling Community Perspectives in Body-Worn Camera Footage Tag Overlap
• Yor-Sarc: A gold-standard dataset for sarcasm detection in a low-resource African language Tag Overlap