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

    Neuronpedia

    APIAssistant AxisNEWCircuit 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. 84676
    Prev
    Next
    INDEX
    Explanations

    to

    np_max-act · gemini-2.0-flash

    This neuron detects mentions of someone or something having ties or connections to a person, group, or place.

    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
     Performs
    -0.07
     BEGIN
    -0.06
    公園
    -0.06
    ุญ
    -0.06
     ПО
    -0.06
    右
    -0.06
     hỏi
    -0.06
    .Bunifu
    -0.06
    委員
    -0.06
     besides
    -0.06
    POSITIVE LOGITS
     associated
    0.07
    EDIA
    0.07
     Brotherhood
    0.06
    ETER
    0.06
     connections
    0.06
     NotImplementedException
    0.06
     cyk
    0.06
    ¨¨
    0.06
    Hardware
    0.06
    Wire
    0.06
    Activations Density 0.020%

    No Known Activations