© 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-16K
    7. 417
    Prev
    Next
    INDEX
    Explanations

    refusal and safety-guideline language where the assistant declines harmful or inappropriate requests, cites ethical/safety policies, and redirects with supportive, non-harmful alternatives.

    oai_token-act-pair · gpt-5Triggered by @irajjmoradi
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-1b-it/resid_post/layer_13_width_16k_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
     var
    0.78
     seven
    0.77
     Six
    0.77
     six
    0.77
     car
    0.77
     secret
    0.76
     K
    0.76
     big
    0.76
     Qu
    0.76
     five
    0.76
    POSITIVE LOGITS
     uphold
    1.08
     responsibly
    0.97
     furtherance
    0.92
     NONINFRINGEMENT
    0.89
    <unused693>
    0.89
    <unused2197>
    0.86
     gewährleisten
    0.86
    䂺
    0.85
     refrained
    0.85
    <unused636>
    0.84
    Activations Density 6.610%

    No Known Activations