© 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. 13998
    Prev
    Next
    INDEX
    Explanations

    pronouns combined with verbs of speech or emotional state

    oai_token-act-pair · gemini-2.0-flash

    characters indicating speech or dialogue

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

    Fiction snippets

    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
     betweenstory
    -0.69
    Према
    -0.59
    لكتر
    -0.57
     craindre
    -0.57
     laude
    -0.57
    quial
    -0.56
     conflit
    -0.56
     daewoo
    -0.55
     morire
    -0.53
     עליו
    -0.53
    POSITIVE LOGITS
    WithIOException
    0.56
     cross
    0.48
     pat
    0.46
     sit
    0.46
    gpl
    0.45
     give
    0.44
    務省
    0.44
     lead
    0.44
     doInBackground
    0.44
     sat
    0.44
    Activations Density 0.739%

    No Known Activations