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. 130644
    Prev
    Next
    INDEX
    Explanations

    skin improvement

    np_max-act · gemini-2.0-flash

    This neuron activates on marketing‐style cosmetic benefit claims—phrases that highlight visible, quantifiable effects (e.g. target, diminish, brighten, visibly, spot repair).

    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
     beers
    -0.07
     logic
    -0.07
    στρο
    -0.06
     Wahl
    -0.06
     tapes
    -0.06
     ORM
    -0.06
    ستی
    -0.06
    belt
    -0.06
    Drivers
    -0.06
    ційна
    -0.06
    POSITIVE LOGITS
    ]=="
    0.07
    "],["
    0.07
     intercept
    0.07
    atown
    0.07
     Dana
    0.06
     ]);
    0.06
     Hollow
    0.06
    .easy
    0.06
     обра
    0.06
     снова
    0.06
    Activations Density 0.031%

    No Known Activations