© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    5. Llama3.1-8B-IT (Instruct)
    7. Resid Post - 131k
    9. 11-RESID-POST-AA
    11. 52293
    Prev
    Next
    INDEX
    Explanations

    witch

    np_max-act · gemini-2.0-flash
    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
    ypad
    -0.08
     removing
    -0.08
    omap
    -0.07
    (undefined
    -0.07
    -food
    -0.07
    พ
    -0.07
     deltas
    -0.07
    .vol
    -0.06
     presumption
    -0.06
    $request
    -0.06
    POSITIVE LOGITS
     witch
    0.13
     Witch
    0.11
     witches
    0.08
    witch
    0.08
    vě
    0.07
    ritch
    0.06
     Pitch
    0.06
    евич
    0.06
    itch
    0.06
    rather
    0.06
    Activations Density 0.003%

    No Known Activations