© 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. 66851
    Prev
    Next
    INDEX
    Explanations

    ghosts and spirits

    np_max-act · gemini-2.0-flash

    This neuron detects references to supernatural, ghostly entities and haunting (mentions of spirits, apparitions, and related phenomena).

    oai_token-act-pair · gpt-5-miniTriggered by @vetterc0
    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
     Epstein
    -0.08
    [column
    -0.07
     asynchronous
    -0.07
     integers
    -0.06
    Discussion
    -0.06
    amarin
    -0.06
     laboratory
    -0.06
    fieldname
    -0.06
    .iterator
    -0.06
     conforms
    -0.06
    POSITIVE LOGITS
     spirits
    0.07
     Portug
    0.06
    reaction
    0.06
    /.
    0.06
     Lightning
    0.06
     dense
    0.06
     dů
    0.06
     частини
    0.06
    ,↵↵
    0.06
    無し�
    0.06
    Activations Density 0.018%

    No Known Activations