© 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. 58009
    Prev
    Next
    INDEX
    Explanations

    start

    np_max-act · gemini-2.0-flash

    The neuron activates on the word “Start,” i.e. it detects the instruction prompt to begin the assistant’s answer.

    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
    maları
    -0.06
    _Selected
    -0.06
    contres
    -0.06
     EACH
    -0.06
    .Th
    -0.06
     přij
    -0.06
    getJSON
    -0.06
    benchmark
    -0.06
    (addr
    -0.06
    iationException
    -0.06
    POSITIVE LOGITS
    agento
    0.07
     Initialize
    0.07
     Він
    0.07
    Start
    0.07
     Мак
    0.07
     nüfus
    0.06
     Vend
    0.06
    ूब
    0.06
    175
    0.06
     Doug
    0.06
    Activations Density 0.002%

    No Known Activations