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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Gemma-2-2B
    3. 24-GEMMASCOPE-TRANSCODER-16K
    4. 195
    Prev
    Next
    INDEX
    Explanations

    long strings of seemingly random characters and numbers, especially within HTML or Javascript code

    oai_token-act-pair · gemini-2.0-flash

    Code and URLs

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2b-pt-transcoders/layer_24/width_16k/average_l0_37
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.24.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
     Z
    -0.22
     (
    -0.20
     Q
    -0.20
     ,
    -0.12
     V
    -0.12
     “
    -0.12
     From
    -0.12
     q
    -0.11
     -
    -0.11
     χ
    -0.11
    POSITIVE LOGITS
     itſelf
    1.27
     myſelf
    1.23
     pleaſure
    1.19
    ſelf
    1.18
     greateſt
    1.16
     purpoſe
    1.15
     Jefus
    1.15
     preſent
    1.13
     Efq
    1.13
     ſtate
    1.12
    Activations Density 20.794%

    No Known Activations