© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    5. Llama3.1-8B-IT
    7. Resid Post - 131k
    9. 11-RESID-POST-AA
    11. 7709
    Prev
    Next
    INDEX
    Explanations

    people

    np_max-act · gemini-2.0-flash

    This neuron consistently lights up on agent-denoting noun endings (e.g. words ending in “-er,” “-ist,” etc.), i.e. terms naming people or roles.

    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
    .pad
    -0.07
     Hamp
    -0.07
    系统
    -0.07
    功能
    -0.06
    Rank
    -0.06
    toBe
    -0.06
    Resistance
    -0.06
     Sergeant
    -0.06
     Blow
    -0.06
     性
    -0.06
    POSITIVE LOGITS
     elites
    0.07
    ＼
    0.06
     GIF
    0.06
    assign
    0.06
     xmlhttp
    0.06
     pres
    0.06
     direccion
    0.06
     Modeling
    0.06
    균
    0.06
     anlat
    0.06
    Activations Density 0.053%

    No Known Activations