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. 125508
    Prev
    Next
    INDEX
    Explanations

    expert

    np_max-act · gemini-2.0-flash

    The neuron activates on words highlighting professional expertise or specialized services (e.g., “expert,” “professional,” “knowledge,” “specialized”).

    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
     pointless
    -0.06
    game
    -0.06
     SIMPLE
    -0.06
    .box
    -0.06
     großen
    -0.06
    озна
    -0.06
     persists
    -0.06
     zz
    -0.06
     सबस
    -0.06
    (plane
    -0.06
    POSITIVE LOGITS
     свидетель
    0.07
     Scholar
    0.07
     sniper
    0.07
     cannabin
    0.06
    _ori
    0.06
     Caucus
    0.06
     verdade
    0.06
     government
    0.06
    โซ
    0.06
     ||=
    0.06
    Activations Density 0.008%

    No Known Activations