© 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. 22-GEMMASCOPE-TRANSCODER-16K
    7. 6380
    Prev
    Next
    INDEX
    Explanations

    words, particularly prefixes, that begin with "out" or "un", and words having to do with cells, plus several seemingly random words.

    oai_token-act-pair · gemini-2.0-flash

    research results

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2b-pt-transcoders/layer_22/width_16k/average_l0_15
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.22.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
    polation
    -0.98
    FTFY
    -0.89
    skosten
    -0.86
    ercises
    -0.85
    unpublished
    -0.85
    lijck
    -0.84
    etheless
    -0.83
    couvrez
    -0.83
    ficulty
    -0.82
    gehalt
    -0.82
    POSITIVE LOGITS
    com
    0.81
    po
    0.80
    form
    0.79
    bu
    0.79
    set
    0.78
    class
    0.77
    post
    0.76
    con
    0.76
    pro
    0.75
    par
    0.74
    Activations Density 9.944%

    No Known Activations