© 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. 30829
    Prev
    Next
    INDEX
    Explanations

    Girl

    np_max-act · gemini-2.0-flash

    The neuron is tuned to pick out speaker‐label tokens (e.g. “NAME_1,” “NAME_2,” the “|” separators, header markers) that mark turns in a dialogue.

    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
     bullshit
    -0.06
     statistically
    -0.06
     logistic
    -0.06
    _PERSON
    -0.06
     ASE
    -0.06
    ascar
    -0.06
     share
    -0.06
    _perc
    -0.06
    asket
    -0.06
     Giz
    -0.06
    POSITIVE LOGITS
     дня
    0.07
     св
    0.07
    оля
    0.07
     طبقه
    0.07
    quierda
    0.06
    ุตสาห
    0.06
     clic
    0.06
     इसस
    0.06
    formData
    0.06
    柔
    0.06
    Activations Density 0.026%

    No Known Activations