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. 81432
    Prev
    Next
    INDEX
    Explanations

    Describing fashion/style

    np_max-act · gemini-2.0-flash

    The neuron focuses on words that describe decorative product features—typically past‐participle or adjective terms highlighting ornate, embellished details.

    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
    castle
    -0.07
    ESC
    -0.06
     EnumerableStream
    -0.06
     Endpoint
    -0.06
     beverages
    -0.06
    Dll
    -0.06
    iotics
    -0.06
    (userid
    -0.06
     )]↵
    -0.06
     then
    -0.06
    POSITIVE LOGITS
    관계
    0.07
    objectManager
    0.06
    ueil
    0.06
     pellet
    0.06
     extravagant
    0.06
    _POWER
    0.06
    aaa
    0.06
    几个
    0.06
    ��
    0.06
    шила
    0.06
    Activations Density 0.039%

    No Known Activations