Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    3. Llama3.1-8B-IT (Instruct)
    4. Resid Post - 131k
    5. 11-RESID-POST-AA
    6. 130653
    Prev
    Next
    INDEX
    Explanations

    punctuation

    np_max-act · gemini-2.0-flash

    The neuron primarily activates on the statement “He/She is a nurse.” (i.e. it detects occurrences of the profession “nurse.”)

    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
     appending
    -0.07
    _movement
    -0.07
    เจ
    -0.06
     confusing
    -0.06
     roaring
    -0.06
     Wars
    -0.06
     Palace
    -0.06
    (IServiceCollection
    -0.06
     Tale
    -0.06
    .App
    -0.06
    POSITIVE LOGITS
    POCH
    0.06
     discharged
    0.06
    land
    0.06
     inconvenient
    0.06
     اضافه
    0.06
     veri
    0.06
     latent
    0.06
    ากาศ
    0.06
    mutation
    0.06
     getId
    0.06
    Activations Density 0.116%

    No Known Activations