© 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. 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. 6215
    Prev
    Next
    INDEX
    Explanations

    guilty

    np_max-act · gemini-2.0-flash

    The neuron is tuned to pick out occurrences of the word “guilty,” especially in the context of guilty pleas.

    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
    vements
    -0.06
    اية
    -0.06
    _LENGTH
    -0.06
    受
    -0.06
     Dread
    -0.06
     Bannon
    -0.06
    ิยม
    -0.06
     Substitute
    -0.06
    ψης
    -0.06
    ерів
    -0.06
    POSITIVE LOGITS
    <y
    0.07
    secondary
    0.07
    ty
    0.07
     inside
    0.07
     CPU
    0.07
     moduleName
    0.07
     screenplay
    0.07
     susceptible
    0.07
    .localStorage
    0.07
    루
    0.07
    Activations Density 0.002%

    No Known Activations