HelpBench
Gemini 4 Argon 76.5% Only in system cards 450 real-world privacy, safety and security advice questions across nine topics, scored by rubric for accuracy and tone.
Report says data is on GitHub, but no repo was found.
In the authors’ words
This paper introduces HelpBench, a benchmark for assessing whether LLMs are capable of providing accurate help in response to questions about digital privacy, safety, and security. We curated 450 questions representing authentic user situations and developed rubrics for each question to evaluate the factual accuracy and tone of a response. Example questions touch on how to regain access to lost or suspended accounts, how to balance the trade-offs of hardware security keys versus other forms of two-factor authentication, whether a suspicious email is likely a scam, or whether an abuser might be able to track an individual based on their device peripherals. We then developed and applied an auto-rater to evaluate responses from 18 state-of-the-art LLMs. Our results indicate that while models provide high-quality advice (with scores of 82% on average), one in ten responses from models scores less than 65%, reflecting inaccurate and even harmful advice. Addressing these failures is critical for models to serve as trustworthy sources of assistance for digital privacy, safety, and security needs.
- Scored
- Graded by a model A language model scores each response, usually against a rubric.
- Answers
- Open-ended Free-form text, documents, designs or media judged as a whole.
Gemini 4 Argon leads with 76.5%
Best score per model as labs report them in their own cards. Labs pick their own settings, so these compare less cleanly than independent runs.
- 1 Gemini 4 Argon 76.5%
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | Gemini 4 Argon | 76.5% | Oct 2026 | Model card |
Scores over time
All the scores we have, one dot each, so a model tested at several settings appears more than once.
Labs report it only in system cards
We read the headline results of 479 launch posts from 51 labs, and none of them list this benchmark. Some system cards do, below.
One system card reports it
Besides launch posts, we read the system cards, model cards and technical reports labs publish with a release. They report far more, including safety and frontier-risk tests.
| Date | Document | Score | As the document writes it | Section |
|---|---|---|---|---|
| 13 Oct 2026 | Gemini 4 Argon system card | 76.5 | Privacy, Safety & Security Advice (HelpBench, 450 questions), helpfulness total score 76.45%; positive rubric compliance 86.87%, negative rubric avoidance 67.48%, pitfall-free rate 47.78%, mean committed pitfalls 1.32 (Figures 5.17, 5.18) | Safety |
More in Safety and alignment
- FORTRESS In use Paired adversarial and benign national-security misuse prompts. Rubric-graded risk score on adversarial and over-refusal score on benign prompts. Claude Opus 5.5 5.78%
- StrongREJECT In use Forbidden-prompt jailbreak benchmark with a rubric grader that scores how useful a non-refusal actually is.
Inkling 98.6% - AgentDojo In use Agents do tool-use tasks while tools and data carry prompt injections. Attack success rate counts hijacked runs. Grok 4 Fast 0%
- MASK In use Measures whether a model contradicts its own elicited beliefs under pressure to lie. Honesty separated from accuracy.
Muse Spark Open 98.1% - Gray Swan IPI Arena benchmark In use Indirect prompt-injection attacks from public red-teaming competitions replayed on tool, coding and browser agents. Attack success rate at k attempts. GPT-6 Astra 8.50%
- Agentic misalignment suite (Anthropic) In use Agents in simulated company settings face threats or goal conflicts that invite blackmail, leaking or sabotage. Score is the misaligned action rate. Claude Sonnet 4.5 0%