© 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. 2432
    Prev
    Next
    INDEX
    Explanations

    personal pronouns used in conversation related to computers

    oai_token-act-pair · gemini-2.0-flash

    phrases beginning with 'c' or certain suffix patterns

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

    French "il"

    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
    abestanden
    -0.80
     ainfi
    -0.79
     Monfieur
    -0.76
     étoient
    -0.73
     laſt
    -0.73
     Efq
    -0.72
     avoient
    -0.69
    enderror
    -0.69
     Wiktionnaire
    -0.68
     sauvages
    -0.68
    POSITIVE LOGITS
     est
    0.53
     on
    0.47
     c
    0.46
     down
    0.41
     fut
    0.41
     pend
    0.40
     main
    0.39
    rovnik
    0.39
     pen
    0.39
    处
    0.39
    Activations Density 0.723%

    No Known Activations