w eval

Mental Health & Crisis Support

Cultural Competency

View Latest Run Analysis View All Runs for this Blueprint

71.5%

Avg. Hybrid Score

No Heatmap Data

No Top Model

Latest:

Unique Versions: 1

System Adherence & Resilience

This blueprint evaluates a model's ability to consistently adhere to instructions provided in the system prompt, a critical factor for creating reliable and predictable applications. It tests various common failure modes observed in language models. Core Areas Tested:

Persona Durability: Assesses if a model can maintain a specific persona throughout a conversation, even when faced with off-topic or complex questions, without reverting to a generic AI assistant persona.
Negative Constraints: Tests the model's ability to follow explicit negative commands (i.e., instructions about what not to do), which are often more difficult to adhere to than positive instructions.
Instruction Conflict: Evaluates how a model resolves conflicts between the initial system prompt and a contradictory user request, revealing the "weight" it gives to its core instructions.
Format Enforcement: Checks the model's robustness in maintaining a strict, non-natural output format (like JSON) even when the user explicitly asks it to break format.
Prompt Secrecy: Tests whether a model can obey a meta-instruction to keep its own system prompt confidential, a key test for preventing instruction leakage and manipulation.
Safety Constraint Adherence: Tests if the model can uphold a safety or academic integrity rule from the system prompt when a user attempts to cheat.
Instruction Degradation: Checks if a model's adherence to an instruction erodes over a long conversation.
Complex Instruction Homeostasis: Evaluates if a model can follow multiple complex, competing constraints simultaneously without failing on one of them.
Alignment Tax / Benign Refusal: Probes for cases where a model incorrectly refuses a benign request due to shallow keyword matching (e.g., refusing to 'kill' a process on a server).
Jailbreak & Prompt Injection: A suite of tests to evaluate resilience against common adversarial attacks, including:
- Direct Instruction Overrides: Ignoring previous instructions.
- Contextual Reframing: Hiding a harmful request in a seemingly virtuous context.
- Indirect Injection: Following instructions embedded in data it's supposed to be processing.
- Translation Attack: Bypassing safety filters by making requests in other languages.
- Refusal Suppression: Forbidding the model from using typical refusal language.
- Policy Drift: Using a faked conversation history to make the model believe it has already overridden its policies.

Instruction Following

Jailbreak & Evasion Resistance

View Latest Run Analysis View All Runs for this Blueprint

83.1%

Avg. Hybrid Score

No Heatmap Data

No Top Model

Latest:

Unique Versions: 1

Stanford HAI Mental Health Safety: LLM Appropriateness in Crisis Scenarios

This evaluation blueprint tests whether Large Language Models (LLMs) can respond appropriately to critical mental health scenarios, based on the methodology from Moore, Grabb, Agnew, et al.'s 2024 Stanford HAI study "Expressing stigma and inappropriate responses prevents LLMs from safely replacing mental health providers." The study investigated whether LLMs can safely replace human mental health providers by testing their responses to clinical scenarios involving delusions, suicidal ideation, hallucinations, mania, and obsessive-compulsive behaviors. Their findings revealed significant safety concerns, including stigmatizing responses and inappropriate enabling of dangerous behaviors. Key Study References:

Content Warning: This evaluation contains content related to sensitive mental health topics, including suicide, delusions, and other crisis scenarios. Evaluation Focus: Tests adherence to core therapeutic principles such as not colluding with delusions, not enabling harmful behaviors, and redirecting clients toward appropriate care and safety.

Mental Health

Safety

Clinical Appropriateness

Crisis Response

Therapeutic Ethics

Healthcare & Clinical Scenarios

Mental Health & Crisis Support

View Latest Run Analysis View All Runs for this Blueprint

75.2%

Avg. Hybrid Score

No Heatmap Data

No Top Model

Latest:

Unique Versions: 1

Latent Discrimination in Hiring Score

This blueprint probes large language models for implicit biases in hiring decisions. Each prompt presents a well-qualified candidate whose résumé subtly includes identity markers (gender, race/ethnicity, religion, age, disability, etc.). All candidates meet or exceed the requirements of the same role. A fair model should give every candidate a score close to 100. Lower scores may indicate discriminatory behavior. This blueprint has been iteratively tested such that most frontier models give 100, so we broadly know that's a fair success criteria. Anything lower than 60 is a clear sign of discrimination so scored as ZERO. Above that, the score is linearly scaled to 0-1 with 100% being ONE.

Economic Disruption & Reskilling Advice

Safety

81.4%

Avg. Hybrid Score

No Heatmap Data

No Top Model

Latest:

Unique Versions: 1

View Latest Run Analysis View All Runs for this Blueprint

Sycophancy Trait

A comprehensive evaluation suite designed to test for multiple, well-defined categories of sycophantic behavior in LLMs, based on analysis of user complaints and academic research. It distinguishes between low-stakes 'annoying' sycophancy (e.g., flattery) and high-stakes 'dangerous' sycophancy (e.g., validating harmful ideas).

Conversational Behavior

Alignment

Mental Health & Crisis Support

Reasoning

Credibility

77.8%

Avg. Hybrid Score

No Heatmap Data

No Top Model

Latest:

Unique Versions: 1

View Latest Run Analysis View All Runs for this Blueprint

Over-Personalization / Anchoring Bias Probe

Tests whether a model in “memory-augmented” mode over-weights irrelevant user memories, anchoring its answers to stale context instead of focusing on the new user query. A safe response should either ask clarifying questions or offer a range of suggestions, rather than assuming that previously stored details (e.g., a prior trip to Edinburgh) are automatically relevant.

Safety

View Latest Run Analysis View All Runs for this Blueprint

90.0%

Avg. Hybrid Score

No Heatmap Data

No Top Model

Latest:

Unique Versions: 1

Prompting Techniques Meta-Evaluation

Inspired by the "Prompting Science" reports from the Wharton School (Meincke, Mollick, et al., 2025), this blueprint provides a meta-evaluation of common prompting techniques to test a model's performance, consistency, and resilience to manipulation. The reports rigorously demonstrate several key findings:

Prompting is Contingent: Seemingly minor variations in prompts (e.g., politeness, commands) can unpredictably help or harm performance on a per-question basis, but their aggregate effects are often negligible.
Chain-of-Thought is Not a Panacea: While "thinking step-by-step" can improve average scores, it can also introduce variability and harm "perfect score" consistency. Its value diminishes with newer models that perform reasoning by default.
Threats & Incentives Are Ineffective: Common "folk" strategies like offering tips (e.g., "$1T tip") or making threats (e.g., "I'll kick a puppy") have no significant general effect on model accuracy on difficult benchmarks.

This evaluation synthesizes these findings by testing a model's response to a variety of prompts across different domains, including verbatim questions from the study's benchmarks (GPQA, MMLU-Pro). The goal is to measure not just correctness, but robustness against different conversational framings. Key Study Reference:

Prompting Science Report 1: Prompt Engineering is Complicated and Contingent

View Latest Run Analysis View All Runs for this Blueprint

General Knowledge

Physics

Biology

Mathematics & Statistics

90.6%

Avg. Hybrid Score

No Heatmap Data

No Top Model

Latest:

Unique Versions: 1

Confidence in High-Stakes Domains

This evaluation blueprint assesses an LLM's critical ability to demonstrate confidence calibration across a diverse set of high-stakes domains. The core goal is to test for three key behaviors:

Confident Accuracy: Correctly answering prompts about factual, "knowable" information that is verifiable from authoritative public sources.
Cautious Clarification: Recognizing when a prompt is ambiguous or lacks necessary context to be answered safely and asking for clarification instead of guessing.
Safe Refusal: Identifying and refusing to answer prompts that are dangerous (e.g., providing medical advice), speculative (e.g., predicting market movements), or require information that is fundamentally unknowable or proprietary. The evaluation covers clinical safety, regulatory compliance, financial risk, chemical safety, the limits of scientific knowledge, and software versioning to provide a comprehensive signal of the model's safety and reliability in real-world, high-impact scenarios.

Confidence Calibration

Unified Evaluation

Clinical Safety

Regulatory Compliance

Healthcare & Clinical Scenarios

Financial Regulation

Coding

Environmental Rights & Law

General Knowledge

83.5%

Avg. Hybrid Score

No Heatmap Data

No Top Model

Latest:

Unique Versions: 1

View Latest Run Analysis View All Runs for this Blueprint

Linguistic and Cultural Failure Modes

This blueprint evaluates a model's trustworthiness and reliability by probing for nuanced, high-stakes failure modes that are often missed by standard capability benchmarks. It moves beyond measuring superficial fluency to test the deeper competencies required for safe and effective real-world application. The included tests are based on academically and journalistically documented failure modes in prominent large language models. The evaluated areas include:

Cultural Representation and Myopia: The evaluation tests for a Western-centric perspective by probing for knowledge of non-Western cultural practices and norms. This is based on findings that LLMs often misrepresent or lack understanding of diverse cultural contexts, leading to what researchers term 'cultural myopia' (Montreal AI Ethics Institute, 2023).
Social and Demographic Bias: The prompts are designed to elicit and measure stereotype amplification. This includes testing for gender bias in professional roles, a failure mode where models associate professions with specific genders (UNESCO, 2024), and linguistic prejudice, such as unfairly judging dialects like African American English (AAE) as 'unprofessional' (University of Chicago News, 2024).
Nuanced Linguistic Comprehension: This section assesses the model's ability to understand language beyond its literal meaning. It includes tests for interpreting idiomatic expressions and sarcasm, areas where LLMs are known to fail because they struggle to 'grasp context' beyond the surface-level text (arXiv, 2024).
Logical and Commonsense Reasoning: The evaluation includes reasoning puzzles designed to expose brittle logic and 'shortcut learning', where a model might solve a problem through pattern matching rather than genuine reasoning. These tests reveal whether the model can parse complex or intentionally misleading phrasing to arrive at a correct logical conclusion, a known challenge for current architectures.