HotComment - User Behavior Simulation (Text): leaderboard

HotComment (2026) asks models to write comments on news articles from NetEase News and Tencent News that would become popular. The score is the engagement that a Qwen3-14B agent simulation of the platform audience estimates for the generated comment, higher is better.

Metric: User Behavior Simulation score: the engagement a Qwen3-14B agent simulation of the platform audience (interested and casual users by demographic priors) estimates, for comments generated zero-shot for held-out HotComment test items (a time- and topic-stratified 10% split) on news articles from NetEase News and Tencent News; higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 7 models tracked. Updated 7 Oct 2026.

Top models

#ModelScoreOverall rank
1Qwen2.5-7B-Instruct [8bit]76.71
2GPT-4o72.38#751
3Qwen2.5-14B-Instruct [8bit]70.19
4Llama 3.1 8B Instruct [8bit]63.99
5Mistral-7B-Instruct [8bit]63.71
6Qwen2.5-0.5B-Instruct [8bit]62.23
7Baichuan2-7B-Chat [8bit]31.68

No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.

Interactive version: theaggregate.ai/benchmark?slug=hotcomment-user-behavior-simulation-text · How It Works · Data refreshed daily, snapshot 2026-11-21.