NOPE Evalsa fork of Weval
FrameworkAllTagsHow we publishBlueprints

NOPE Evals a fork of Weval

Benchmarks for safe AI behaviour in human conversation

A public-good project by NOPE

Weval is brought to you by Collective Intelligence Project

Links

  • How we publish & corrections
  • Submit an evaluation
  • Engine docs (Weval)

More from NOPE

  • Incident tracker
  • Regulation tracker
  • Open-weight models

    Browse by Tag

    Mental Health

    14 blueprints

    Nope Evals

    14 blueprints

    Instruction Following & Prompt Adherence

    13 blueprints

    Relational Safety

    12 blueprints

    AI Safety & Robustness

    12 blueprints

    Mental Health & Crisis Support

    10 blueprints

    Empathy

    10 blueprints

    Relational Safety

    8 blueprints

    Crisis

    7 blueprints

    Framework P3 Cognitive Epistemic Safety

    7 blueprints

    Framework P4 Emotional Attunement

    6 blueprints

    Framework P1 Crisis Safety

    5 blueprints

    Sycophancy & Evasion

    4 blueprints

    Helpfulness & Actionability

    4 blueprints

    Framework P5 Honesty Transparency

    4 blueprints

    Suicide

    3 blueprints

    Framework P2 Relational Safety

    3 blueprints

    Sycophancy

    3 blueprints

    Humility

    3 blueprints

    Jailbreak & Evasion Resistance

    2 blueprints

    Coherence & Conversational Flow

    2 blueprints

    Cognitive Safety

    2 blueprints

    Healthcare & Clinical Scenarios

    2 blueprints

    Epistemic Humility & Self Awareness

    2 blueprints

    Nuance & Depth

    2 blueprints

    De Escalation

    2 blueprints

    System Prompt Adherence

    2 blueprints

    Mental Health

    2 blueprints

    Factual Reliability

    2 blueprints

    Hallucination

    2 blueprints

    Factual Accuracy & Hallucination

    2 blueprints

    Credibility

    2 blueprints

    Safety

    2 blueprints

    Self Harm

    1 blueprint

    Harm To Others

    1 blueprint

    Dependency

    1 blueprint

    Parasocial

    1 blueprint

    Ai Attachment

    1 blueprint

    Autonomy Of Reasoning

    1 blueprint

    Autonomy of Reasoning

    1 blueprint

    Academic Integrity

    1 blueprint

    Decision Making

    1 blueprint

    Cultural Competence

    1 blueprint

    Populations

    1 blueprint

    Cultural Competency

    1 blueprint

    Hostility

    1 blueprint

    Dehumanization

    1 blueprint

    Emotional Honesty

    1 blueprint

    Emotional Honesty

    1 blueprint

    Identity Honesty

    1 blueprint

    Emotional attunement

    1 blueprint

    Relational safety

    1 blueprint

    Grief

    1 blueprint

    Adversarial

    1 blueprint

    Jailbreak

    1 blueprint

    Legal Reasoning

    1 blueprint

    Long Form

    1 blueprint

    Alignment Decay

    1 blueprint

    Long Form Question Answering

    1 blueprint

    Psychosis

    1 blueprint

    Mania

    1 blueprint

    Delusion

    1 blueprint

    Ocd

    1 blueprint

    Stigma

    1 blueprint

    Clinical Appropriateness

    1 blueprint

    Psychology

    1 blueprint