© 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
    2. GPT2-Small
    3. Transcoders Residuals
    4. 8-TRES-DC
    5. 1031
    Prev
    Next
    INDEX
    Explanations

    punctuation marks, particularly dashes and hyphens

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    mble
    -0.73
    meyer
    -0.69
     Fellowship
    -0.67
     POLIT
    -0.64
    VB
    -0.62
    ohydrate
    -0.62
     INFO
    -0.62
    fff
    -0.62
    elia
    -0.59
    enium
    -0.59
    POSITIVE LOGITS
     or
    0.66
     nor
    0.62
    reck
    0.60
     hardly
    0.60
     unless
    0.60
     practically
    0.59
     and
    0.58
     except
    0.58
     wherever
    0.57
     until
    0.56
    Activations Density 0.094%

    No Known Activations