Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. GPT2-Small
    3. Transcoders Residuals
    4. 8-TRES-DC
    5. 226
    Prev
    Next
    INDEX
    Explanations

    instances of the word "maintain" and its variations, indicating a focus on the concept of preservation or ongoing support

    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
    sonian
    -0.74
    agate
    -0.73
    SOURCE
    -0.70
    olini
    -0.69
    illion
    -0.63
    here
    -0.61
    ividual
    -0.61
    VIDIA
    -0.60
    TAIN
    -0.60
    andel
    -0.60
    POSITIVE LOGITS
     anonymity
    0.86
    oxide
    0.79
     vigilance
    0.79
     calm
    0.75
     confidence
    0.74
     proficiency
    0.73
     sanity
    0.72
     dignity
    0.71
     purity
    0.70
     impartial
    0.69
    Activations Density 0.038%

    No Known Activations