© 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. 3232
    Prev
    Next
    INDEX
    Explanations

    the phrase "going to"

    oai_token-act-pair · gemini-2.0-flash

    say "to"

    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
    goes
    -0.68
    antMatchers
    -0.64
     goes
    -0.63
    ReusableCell
    -0.59
    went
    -0.59
     went
    -0.59
     gik
    -0.57
     IBOutlet
    -0.56
     Theſe
    -0.56
    omenclature
    -0.55
    POSITIVE LOGITS
     Roskov
    0.92
    ArrowToggle
    0.52
    TintMode
    0.52
     qualify
    0.51
    AutoField
    0.51
     Nadu
    0.50
    localize
    0.48
     ratify
    0.47
     facie
    0.46
    devise
    0.46
    Activations Density 0.525%

    No Known Activations