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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & 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. 54891
    Prev
    Next
    INDEX
    Explanations

    Locations

    np_max-act · gemini-2.0-flash

    The neuron specifically picks out numeric and punctuation tokens used in address or location 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
     McGu
    -0.07
     edilmiştir
    -0.06
     البي
    -0.06
     ◎
    -0.06
     etmiş
    -0.06
     LZ
    -0.06
    ňují
    -0.06
     Zem
    -0.06
     Hồ
    -0.06
     회원
    -0.06
    POSITIVE LOGITS
    //================================================================
    0.09
    ۴
    0.07
    divider
    0.07
     Garc
    0.07
     Ensure
    0.07
    Selected
    0.07
     actually
    0.06
     Depends
    0.06
    T
    0.06
    =result
    0.06
    Activations Density 0.019%

    No Known Activations