© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-1B-IT
    5. 13-GEMMASCOPE-2-RES-65K
    7. 440
    Prev
    Next
    INDEX
    Explanations

    phrases where the model acknowledges a user's emotional distress, refuses a harmful request, and offers supportive alternatives or resources.

    oai_token-act-pair · claude-4-5-haikuTriggered by @hkale7
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-1b-it/resid_post/layer_13_width_65k_l0_medium
    Prompts (Dashboard)
    238,145 prompts, 512 tokens each
    Dataset (Dashboard)
    lmsys + oasst1
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     особенно
    0.63
     velocity
    0.62
    L
    0.61
     Y
    0.61
    especially
    0.58
     taxa
    0.58
    velocity
    0.58
     particularly
    0.58
     L
    0.57
    Y
    0.57
    POSITIVE LOGITS
     alternatives
    1.09
     healthier
    1.07
     alternative
    1.03
     Alternatives
    0.99
     Alternative
    0.99
     alternatif
    0.99
     safer
    0.99
     instead
    0.96
     Instead
    0.96
    代替
    0.95
    Activations Density 5.801%

    No Known Activations