© 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. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    5. Llama3.1-8B-IT
    7. Resid Post - 131k
    9. 11-RESID-POST-AA
    11. 58800
    Prev
    Next
    INDEX
    Explanations

    etc

    np_max-act · gemini-2.0-flash

    The neuron spotlights mentions of system configuration file paths, specifically occurrences of “etc” (as in “/etc/…”).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    andyrdt/saes-llama-3.1-8b-instruct/resid_post_layer_11/trainer_1
    Dataset (Dashboard)
    Various
    Features
    131,072
    Data Type
    float32
    Hook Name
    blocks.11.hook_resid_post
    Architecture
    standard
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Mour
    -0.07
    adors
    -0.07
     imperial
    -0.07
     Superman
    -0.06
     nim
    -0.06
    166
    -0.06
     visions
    -0.06
     Gardner
    -0.06
    .snap
    -0.06
     Quinn
    -0.06
    POSITIVE LOGITS
    /etc
    0.09
    etc
    0.09
    =settings
    0.07
    çı
    0.07
     отверсти
    0.07
    (REG
    0.07
    TC
    0.06
    特别
    0.06
    ความค
    0.06
    =options
    0.06
    Activations Density 0.001%

    No Known Activations