© 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. 2-GEMMASCOPE-TRANSCODER-16K
    7. 10454
    Prev
    Next
    INDEX
    Explanations

    strings of seemingly random characters and numbers including dashes, underscores, and colons, sometimes in code or dates

    oai_token-act-pair · gemini-2.0-flash

    Dates and identifiers

    np_max-act · gemini-2.0-flash

    _ Method used: 1 — because tokens are separators/punctuation

    np_max-act-logits · gpt-5-miniTriggered by @chenshw0109
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2b-pt-transcoders/layer_2/width_16k/average_l0_49
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.2.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
    Xna
    -0.68
    fromnode
    -0.66
    økt
    -0.61
     Wikimedijinoj
    -0.59
    verifyException
    -0.58
    μβ
    -0.57
     HasFactory
    -0.57
    )}</
    -0.56
    '])
    -0.56
     Waray
    -0.55
    POSITIVE LOGITS
    BeginContext
    0.76
    AsUp
    0.55
     oxigen
    0.54
    Transcription
    0.53
    務省
    0.49
    acek
    0.49
     Cielo
    0.47
     SwitchCompat
    0.46
    toare
    0.44
    Nederlands
    0.44
    Activations Density 1.252%

    No Known Activations