© 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
    3. Gemma-2-2B
    5. 22-GEMMASCOPE-TRANSCODER-16K
    7. 10933
    Prev
    Next
    INDEX
    Explanations

    tokens that often occur at the beginning of clauses in arguments and explanations.

    oai_token-act-pair · gemini-2.0-flash

    As

    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
    StoryboardSegue
    -0.61
     TimeUnit
    -0.57
    WriteBarrier
    -0.56
     gourd
    -0.52
    requireNonNull
    -0.50
    
    -0.50
    neté
    -0.50
     NSCoder
    -0.50
     htons
    -0.49
    ngrx
    -0.49
    POSITIVE LOGITS
     is
    1.81
     was
    1.34
     has
    1.16
     isn
    1.07
     seems
    1.07
     doesn
    0.86
     wasn
    0.86
     does
    0.83
     appears
    0.82
     isnt
    0.82
    Activations Density 5.683%

    No Known Activations