© 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. 21-GEMMASCOPE-TRANSCODER-16K
    7. 5475
    Prev
    Next
    INDEX
    Explanations

    words and phrases related to speculations about the future and plans

    oai_token-act-pair · gemini-2.0-flash

    variations of the word "hold"

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

    future tense

    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
    willi
    -0.43
    瓶
    -0.43
     DESCRIP
    -0.42
     richtet
    -0.40
     Finding
    -0.39
    Parameterized
    -0.38
    arrer
    -0.38
    zade
    -0.38
     دنبال
    -0.37
     nhiêu
    -0.37
    POSITIVE LOGITS
     hold
    1.27
     holds
    1.22
    hold
    1.10
     HOLD
    1.06
     Hold
    1.05
     holding
    1.05
    Hold
    1.01
    holds
    0.99
     Holds
    0.99
     Holding
    0.96
    Activations Density 1.515%

    No Known Activations