© 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. GPT2-Small
    5. Transcoders Residuals
    7. 8-TRES-DC
    9. 2530
    Prev
    Next
    INDEX
    Explanations

    instances of the word "improve" and its variants in various contexts

    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
    DragonMagazine
    -1.04
    ories
    -0.80
    rete
    -0.76
    pload
    -0.75
    rys
    -0.73
    lyak
    -0.72
    borg
    -0.68
    子
    -0.67
    orie
    -0.67
    kt
    -0.66
    POSITIVE LOGITS
     accuracy
    0.99
     upon
    0.75
     coordination
    0.73
     decency
    0.70
     morale
    0.69
     localization
    0.69
     fidelity
    0.68
     efficiency
    0.68
     literacy
    0.68
     accur
    0.67
    Activations Density 0.062%

    No Known Activations