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

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-2B
    5. 21-GEMMASCOPE-TRANSCODER-16K
    7. 7205
    Prev
    Next
    INDEX
    Explanations

    sentences describing situations where people are controlled or restricted

    oai_token-act-pair · gemini-2.0-flash

    words related to restrictions and control

    np_token-act-pair-logits · gpt-4o-mini

    rules and restrictions

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2b-pt-transcoders/layer_21/width_16k/average_l0_13
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.21.ln2.hook_normalized
    Architecture
    jumprelu_transcoder
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     broken
    -0.59
    broken
    -0.59
    Broken
    -0.59
     chimiques
    -0.56
     violated
    -0.55
     laajenta
    -0.55
     Broken
    -0.54
     applicazione
    -0.53
     colorés
    -0.52
     remplir
    -0.52
    POSITIVE LOGITS
     governed
    0.68
     שוליים
    0.64
     prevented
    0.64
     blocked
    0.62
     allowed
    0.62
    ьаж
    0.62
     prohibited
    0.62
     barred
    0.60
     Allowed
    0.59
     preceded
    0.59
    Activations Density 3.213%

    No Known Activations