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. 130382
    Prev
    Next
    INDEX
    Explanations

    reunited

    np_max-act · gemini-2.0-flash

    The neuron responds to the model’s internal markup and control tokens (e.g. header IDs, begin/end text tags) rather than actual words.

    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
     punitive
    -0.07
    άζ
    -0.07
    western
    -0.07
     baff
    -0.06
    者の
    -0.06
    imators
    -0.06
     scratched
    -0.06
    �
    -0.06
    -fetch
    -0.06
     academia
    -0.06
    POSITIVE LOGITS
    Pairs
    0.06
     Lie
    0.06
    Fc
    0.06
    403
    0.06
    elling
    0.06
    lbl
    0.06
    jj
    0.06
     bicycles
    0.05
    斗
    0.05
    estate
    0.05
    Activations Density 0.019%

    No Known Activations