© 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. 12190
    Prev
    Next
    INDEX
    Explanations

    multiple languages

    np_max-act · gemini-2.0-flash

    This neuron responds whenever the model is generating English (Latin-script) words—its activations go up on English tokens and punctuation but stay off on non-English text.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8

    phrases related to the process and components involved in a system or workflow.

    oai_token-act-pair · gpt-4o-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
    Elem
    -0.07
    g
    -0.06
     Kinh
    -0.06
     Onun
    -0.06
     fopen
    -0.06
    ihu
    -0.06
     Fr
    -0.06
     Aim
    -0.06
    Six
    -0.06
     Smarty
    -0.06
    POSITIVE LOGITS
     {{↵
    0.07
    [val
    0.07
    \b
    0.06
     obligation
    0.06
     kla
    0.06
    izers
    0.06
     még
    0.06
    ียม
    0.06
     ""){↵
    0.06
     electronic
    0.06
    Activations Density 0.223%

    No Known Activations