© 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. 0-GEMMASCOPE-TRANSCODER-16K
    7. 3772
    Prev
    Next
    INDEX
    Explanations

    colons followed by whitespace

    oai_token-act-pair · gemini-2.0-flash

    code/memory addresses

    np_max-act · gemini-2.0-flash

    ▁ Method used: 1 — reason: identical MAX_ACTIVATING_TOKENS (spacer) — ▁

    np_max-act-logits · gpt-5-miniTriggered by @chenshw0109
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2b-pt-transcoders/layer_0/width_16k/average_l0_76
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.0.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
     
    -3.61
     the
    -2.14
     a
    -1.99
    ↵↵
    -1.95
    <bos>
    -1.87
     I
    -1.84
     in
    -1.84
      
    -1.84
     A
    -1.77
     (
    -1.74
    POSITIVE LOGITS
     Efq
    3.41
     Monfieur
    3.33
     myſelf
    3.31
     Majefty
    3.30
     ―――――
    3.23
     Reſ
    3.23
     Anſ
    3.20
     Houſe
    3.16
     Jefus
    3.09
     itſelf
    3.09
    Activations Density 3.092%

    No Known Activations