© 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. 56358
    Prev
    Next
    INDEX
    Explanations

    Sports action/wrestling

    np_max-act · gemini-2.0-flash

    This neuron selectively responds to dynamic action verbs in play-by-play or ring commentary that describe participants’ movements or staged physical actions (e.g. “congratulated,” “rushed out,” “bring(s) out,” “hold back”).

    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
     commanding
    -0.07
    -й
    -0.06
    Storyboard
    -0.06
     �
    -0.06
    WebService
    -0.06
    .logo
    -0.06
    whel
    -0.06
     Silicon
    -0.06
    Endpoints
    -0.06
    _province
    -0.06
    POSITIVE LOGITS
    土地
    0.07
    断
    0.06
    可是
    0.06
    -panel
    0.06
    -G
    0.06
     scandals
    0.06
     decorate
    0.06
    $('#
    0.06
    $('
    0.06
    П
    0.06
    Activations Density 0.005%

    No Known Activations