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

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI 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. 57765
    Prev
    Next
    INDEX
    Explanations

    Math and science

    np_max-act · gemini-2.0-flash

    The neuron is strongly activated by the character sequence “ads” (as in “adsorption” or “ads”) within words.

    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
    Interestingly
    -0.06
     rents
    -0.06
     Чем
    -0.06
     것이다
    -0.06
     haute
    -0.06
     území
    -0.06
    Creation
    -0.06
     dès
    -0.06
     rave
    -0.06
    ered
    -0.06
    POSITIVE LOGITS
    connections
    0.07
    exchange
    0.06
    .INPUT
    0.06
    ikit
    0.06
    .full
    0.06
    �
    0.06
     trader
    0.06
    )."
    0.06
     Cir
    0.06
    -items
    0.06
    Activations Density 0.167%

    No Known Activations