© 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 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. 38627
    Prev
    Next
    INDEX
    Explanations

    emotions and opinions

    np_max-act · gemini-2.0-flash

    The neuron selectively activates on vivid, atmosphere‐setting words—those descriptive terms that evoke mood or sensory impressions (e.g. “mystique,” “magical,” “hidden,” “wonder,” 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
    _',
    -0.07
     removed
    -0.07
     Doesn
    -0.07
     ball
    -0.06
    $
    -0.06
    ('+
    -0.06
    rstrip
    -0.06
    ΙΟΥ
    -0.06
    .screen
    -0.06
    Normals
    -0.06
    POSITIVE LOGITS
    ック
    0.07
     Mou
    0.06
     kontakt
    0.06
     unlikely
    0.06
     incontri
    0.06
     países
    0.06
     literal
    0.06
    Ham
    0.06
    rase
    0.06
     buoy
    0.06
    Activations Density 0.075%

    No Known Activations