Research map / Large language models / Retrieval and knowledge
OmniBench: A Comprehensive Benchmark Integrating Real-World, Time-sensitive, and Multi-Hop Questions with a Multi-Dimensional Hybrid Evaluation Framework
Wenjie Wang, Yufeng Jiang, Ge Sun, Chenghang Dong, Zheng Jun, Li Mengjie, Lixin Chen, Huan Wang, Haoyu Wang, Bin Chen
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 changed | Where | Measured | Effect |
|---|---|---|---|
| enable deep thinking mode (with internet) | Doubao · OmniBench | Overall score: 65.86 -> 84.62 | Improved |
| enable deep thinking mode (with internet) | DeepSeek · OmniBench | Overall score: 77.659 -> 86.66 | Improved |
| enable deep thinking mode (with internet) | Yuanbao · OmniBench | Overall score: 80.19 -> 83.82 | Mixed |
| enable deep thinking mode (with internet) | Kimi · OmniBench | Overall score: 65.629 -> 65.15 | No clear effect |
| enable deep thinking mode | Doubao · OmniBench | Insight: 48.48 -> 86.80 | Improved |
| enable deep thinking mode | Doubao · OmniBench | Factuality: 74.02 -> 69.54 | Hurt |
| enable deep thinking mode | DeepSeek · OmniBench | Factuality: 69.33 -> 68.90 | Hurt |
| longer responses | AI applications in deep thinking mode · OmniBench | Advanced metrics | Improved |
| longer responses | AI applications · OmniBench | Factuality | Hurt |
| multi-hop questions vs open-ended | DeepSeek · OmniBench | Dataset score: 88.68 -> 76.94 | Hurt |
| time-sensitive questions vs open-ended | OpenAI o4-mini · OmniBench | Dataset score: 61.42 -> 38.30 | Hurt |
Closest papers by meaning
- MDK12-Bench: A Multi-Discipline Benchmark for Evaluating Reasoning in Multimodal Large Language Models (AAAI 2026 · similarity 0.87)
- InfiBench: Evaluating the Question-Answering Capabilities of Code Large Language Models (NeurIPS 2024 · similarity 0.87)
- A Semantic-Sampling Framework for Evaluating Calibration in Open-Ended Question Answering (NeurIPS 2026 · similarity 0.87)
- WildBench: Benchmarking LLMs with Challenging Tasks from Real Users in the Wild (ICLR 2025 · similarity 0.87)
- CofCA: A STEP-WISE Counterfactual Multi-hop QA benchmark (ICLR 2025 · similarity 0.87)
- CruxBench: A Benchmark of Information Discovery (NeurIPS 2026 · similarity 0.87)
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.
