© 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. 108878
    Prev
    Next
    INDEX
    Explanations

    Questions and answers

    np_max-act · gemini-2.0-flash

    This neuron detects instruction-giving language, particularly second-person imperatives and directive phrases (e.g. “you,” “should,” “generate,” “do not”).

    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
     Guth
    -0.06
     defaultProps
    -0.06
     tịch
    -0.06
     mesh
    -0.06
     програ
    -0.06
    pole
    -0.06
    .feedback
    -0.06
     software
    -0.06
     يع
    -0.06
    чи
    -0.06
    POSITIVE LOGITS
    /D
    0.07
     dieses
    0.07
    .isDebugEnabled
    0.07
    Die
    0.07
    Increases
    0.06
     (=
    0.06
    (dot
    0.06
    _Space
    0.06
    )".
    0.06
    (G
    0.06
    Activations Density 0.017%

    No Known Activations