AI Paper Atlas

Research map / Large language models / Agents, code and math

SoMe: A Realistic Benchmark for LLM-based Social Media Agents

Dizhan Xue, Jing Cui, Shengsheng Qian, Chuanrui Hu, Changsheng Xu

AAAI 2026 · Large language models › Agents, code and math · approach: Agent instruction-following benchmarks · 2 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
scale model size from Qwen3-14B to Qwen3-32BQwen3 agent · SoMe average scoreaverage score: 52.82 -> 53.98Improved
scale model size from Qwen3-8B to Qwen3-14BQwen3 agent · SoMe average scoreaverage score: 50.89 -> 52.82Improved
scale from Llama-3.1-8B to Llama-3.3-70BLlama agent · SoMe average scoreaverage score: 31.65 -> 48.34Improved
use reasoning-distilled DeepSeek-R1-0528-Qwen3-8B instead of Qwen3-8B8B agent · SoMe average scoreaverage score: 50.89 -> 31.87Hurt
use larger GLM-4-32B instead of GLM-4-9BGLM-4 agent · SoMe average scoreaverage score: 37.51 -> 38.86Improved
larger model size reduces tool-response hallucinationKimi-K2-Instruct (1T) and DeepSeek-V3 (671B) · SoMe Real-time Event Detection hallucination testtool response hallucination rate: 7% and 13%No clear effect
higher task completion rate (tool-calling ability) goes with better scoresGemini-2.5-Flash, DeepSeek-V3, Qwen3-32B, Qwen3-14B · SoMeaverage score: TCR 96.61 to 98.60Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy