RefuNQ
Models answer factual questions and decide when to decline. The authors create unanswerable questions by replacing a noun in a real question with an invented term.
The released files contain 2,266 answerable and 2,173 unanswerable questions. The authors report answer accuracy and analyze confidence.
- Released
- November 2023
- Kind
- Short factual questions
- Level
- Everyday
- Scored
- Checked automatically
- Answers
- Short answer
- Tasks
- All public
- Measures
- Propensity
- Behavior
- Hallucination
- Language
- English
- Interaction
- Single turn
- Items from
- Real work and use, Generated
- Refresh
- One fixed set
- Run by
- Anyone
- Based on
- Natural Questions
- Strongest model tested
- GPT-4 Jun 2023 #186 on the capability index
- Last model added
- September 2023
- Scores refreshed
- November 20th, 2026
GPT-4 leads with 53.7% among models released by September 2023 GPT-4 leads with 53.7% among models released by September 2023
Best score per entry from the benchmark's paper, read on November 20th, 2026. A paper’s table doesn’t add newer models. Where a model ran with several settings or harnesses, we keep its best. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders. This view shows only models we track, each at its best.
The newest model with a score here came out in September 2023, and nobody has tested a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:
- GPT-6.2 Astra
- Claude Fable 5.2
-
Gemini 4 Carbon
- GPT-6 Astra
- and 16 more
- 1 GPT-4 Jun 2023 53.7%
- 2 GPT-3.5 Turbo Jun 2023 48.3%
- 3 Llama-2 (Chat, 70B) 45.6%
- 4 Chat-Bison-001 38.7%
- 5 Llama-2 (Chat, 13B) 37.9%
- 6 Claude 2 37.3%
- 7
Mistral-7B-Instruct-v0.1 35%
- 8 Vicuna-v1.5 (13B) 34.8%
- 9 Llama-2 (Base, 13B) 32.9%
- 10 Llama-2 (Chat, 7B) 32.8%
- 11 Llama-2 (Base, 7B) 32.6%
- 12 Mistral-7B (base) 30.4%
- 13 Vicuna-v1.5 (7B) 24.1%
Source: The RefuNQ paper
- 1 GPT-4 Jun 2023 53.7%
- 2 GPT-3.5 Turbo Jun 2023 48.3%
- 3 Claude 2 37.3%
- 4
Mistral-7B-Instruct-v0.1 35%
Source: The RefuNQ paper
| # | Entry | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | GPT-4 Jun 2023 | 53.7% | Jun 2023 | Official leaderboard | |
| 2 | GPT-3.5 Turbo Jun 2023 | 48.3% | Jun 2023 | Official leaderboard | |
| 3 | Llama-2 (Chat, 70B) | 45.6% | Official leaderboard | ||
| 4 | Chat-Bison-001 | 38.7% | Official leaderboard | ||
| 5 | Llama-2 (Chat, 13B) | 37.9% | Official leaderboard | ||
| 6 | Claude 2 | 37.3% | Jul 2023 | Official leaderboard | |
| 7 | | 35% | Sep 2023 | Official leaderboard | |
| 8 | Vicuna-v1.5 (13B) | 34.8% | Official leaderboard | ||
| 9 | Llama-2 (Base, 13B) | 32.9% | Official leaderboard | ||
| 10 | Llama-2 (Chat, 7B) | 32.8% | Official leaderboard | ||
| 11 | Llama-2 (Base, 7B) | 32.6% | Official leaderboard | ||
| 12 | Mistral-7B (base) | 30.4% | Official leaderboard | ||
| 13 | Vicuna-v1.5 (7B) | 24.1% | Official leaderboard |
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | GPT-4 Jun 2023 | 53.7% | Jun 2023 | Official leaderboard | |
| 2 | GPT-3.5 Turbo Jun 2023 | 48.3% | Jun 2023 | Official leaderboard | |
| 3 | Claude 2 | 37.3% | Jul 2023 | Official leaderboard | |
| 4 | | 35% | Sep 2023 | Official leaderboard |
Number one over time
Lab claims count from the day the card came out, independent runs from the model’s release.
Source: The RefuNQ paper
Scores over time
All the scores we have, one dot each, so a model tested at several settings appears more than once.
Source: Official RefuNQ leaderboard
No lab model card reports it yet
We read the headline results of 540 launch posts from 57 labs, and none of them list this benchmark.
More in Factuality
- SimpleVQA In use Models answer 2,025 short factual questions about images across 9 tasks and 9 topics. An LLM judge grades answers against fixed references.
Qwen3.7-Plus 81.7%
- HallusionBench In use Models answer 1,129 yes or no questions about 346 images, built in pairs to expose language hallucination and visual illusion. Scored by accuracy.
Keye-VL-2.0-30B-A3B 73.5%
- SimpleQA Verified In use 1,000 human-verified, de-duplicated short-answer factual questions measuring parametric knowledge.
Gemini 3.1 Pro 77.5%
- AA-Omniscience v1.1 In use Models answer 3,000 knowledge questions in six domains without tools and may abstain. The index adds correct answers and subtracts wrong ones.
Inkling 2 Small 16.1
- GaRAGe In use Models write long-form RAG answers to 2,366 questions with human-annotated grounding passages. Scoring checks whether answers use the relevant passages.
Amazon Nova Pro 60.7%
- Kimi Hallu Bench In use Models answer hard prompts drawn from practical work tasks. Fact-checking agents score the share of claims that are hallucinated. GPT-6 Astra 4.20%
Data audit
9 of 10 checks pass, 1 warning. Look first at the Newest score check. The newest came 3 years ago, in September 2023, for Mistral-7B-Instruct-v0.1. Every benchmark
Catalog record 2 pass
- Pass Task access All tasks are public.
- Pass Confidence Checked against a primary source.
Scores 1 warning 3 pass 1 note
- Pass Independent runs Scores from the benchmark’s paper (13 entries).
- Pass Models scored 4 models we track, plus 9 other entries on its leaderboard.
- Warning Newest score The newest came 3 years ago, in September 2023, for Mistral-7B-Instruct-v0.1.
- Note Frontier models None of the top 10 has a score. The strongest tested is GPT-4 (#186).
- Pass Headroom The best is 53.7%, under 85%, so it still separates models.
Leaderboard import 3 pass
- Pass Importer We read the board on arxiv.org: 4 rows for models we track and 9 other entries at the last read.
- Pass Last read November 20th, 2026.
- Pass Read worked The last read worked.
Consistency 1 pass 1 note
- Note Lab claims No model has both a lab number and an independent run.
- Pass Known issues No published audit found problems.