© 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. 24-GEMMASCOPE-TRANSCODER-16K
    7. 14533
    Prev
    Next
    INDEX
    Explanations

    semicolons, parentheses, and other characters commonly used in programming languages

    oai_token-act-pair · gemini-2.0-flash

    code punctuation

    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
     tromper
    -0.82
     enfans
    -0.81
    ientôt
    -0.79
     étoient
    -0.77
     refroid
    -0.77
     stället
    -0.76
    Còn
    -0.76
     étroite
    -0.74
     före
    -0.72
     conséquence
    -0.71
    POSITIVE LOGITS
     Viceroy
    0.39
     Upan
    0.35
     EEU
    0.35
     Asiatic
    0.34
     Chaldean
    0.34
     Hopf
    0.33
     Huguen
    0.33
     handker
    0.32
     Persians
    0.32
     Andean
    0.32
    Activations Density 4.214%

    No Known Activations