© 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. 12894
    Prev
    Next
    INDEX
    Explanations

    the word "improvement" and words modified by it

    oai_token-act-pair · gemini-2.0-flash

    improvement

    np_max-act · gemini-2.0-flash

    improvement Method used: 1 — because every top activating token is the same word, "improvement"

    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
    <bos>
    -1.45
     linkovi
    -0.94
     Paglinawan
    -0.91
    DebuggerNonUser
    -0.84
    SequentialGroup
    -0.73
     թվական
    -0.72
    хьтан
    -0.69
    ActionCreators
    -0.68
    Rè
    -0.68
    ImageContext
    -0.68
    POSITIVE LOGITS
     tranſ
    0.93
     Diſ
    0.91
     pleaſure
    0.91
     Reſ
    0.91
     purpoſe
    0.91
     houſe
    0.90
     Eſ
    0.88
     Inſ
    0.87
     reaſon
    0.86
     Conſ
    0.85
    Activations Density 2.042%

    No Known Activations