© 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
    2. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    3. Llama3.1-8B-IT
    4. Resid Post - 131k
    5. 11-RESID-POST-AA
    6. 30016
    Prev
    Next
    INDEX
    Explanations

    ordinal position

    np_max-act · gemini-2.0-flash

    The neuron activates on ordinal indicators (e.g. “first,” “second,” “third,” 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
     Persons
    -0.07
    _imp
    -0.07
    engage
    -0.07
    weeney
    -0.07
    ьер
    -0.06
    565
    -0.06
     intervention
    -0.06
    ladık
    -0.06
     ther
    -0.06
    ">
    ↵
    ↵
    -0.06
    POSITIVE LOGITS
    (targetEntity
    0.07
     مار
    0.07
     Cena
    0.06
     плат
    0.06
    (Label
    0.06
    ской
    0.06
     지방
    0.06
     coalition
    0.06
     digs
    0.06
     ViewController
    0.06
    Activations Density 0.089%

    No Known Activations