© 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 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. 57318
    Prev
    Next
    INDEX
    Explanations

    First-person narrative

    np_max-act · gemini-2.0-flash

    the main thing this neuron does is find imperative or command-style verbs—i.e. call-to-action instructions.

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

    informational content related to businesses and services.

    oai_token-act-pair · gpt-4o-miniTriggered by @xinyanhu8

    instances of first-person self-reference (I/my/me) or personal, speaker-centered language.

    oai_token-act-pair · gpt-5-miniTriggered by @vetterc0
    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
     we
    -0.11
    We
    -0.10
     We
    -0.09
     us
    -0.09
    Our
    -0.08
    áme
    -0.07
     our
    -0.07
     Our
    -0.07
    (Py
    -0.07
    Come
    -0.07
    POSITIVE LOGITS
     فرودگاه
    0.08
     약
    0.07
     peaks
    0.06
    getDoctrine
    0.06
    ηγ
    0.06
     etk
    0.06
     relief
    0.06
    _engine
    0.06
     llegar
    0.06
     sexkontakte
    0.06
    Activations Density 0.244%

    No Known Activations