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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    5. Llama3.1-8B-IT (Instruct)
    7. Resid Post - 131k
    9. 11-RESID-POST-AA
    11. 92598
    Prev
    Next
    INDEX
    Explanations

    Negation/something not happening

    np_max-act · gemini-2.0-flash

    This neuron detects first‐person, self‐reflective phrases—instances where the writer refers to themselves (“I”, “me”, “hadn't thought of”, “before”, etc.) and expresses personal realization or surprise.

    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
     senin
    -0.07
     Naz
    -0.07
    COMPARE
    -0.07
    Uni
    -0.06
     gerektir
    -0.06
    YLE
    -0.06
    dığını
    -0.06
    mpl
    -0.06
     gv
    -0.06
    ’dan
    -0.06
    POSITIVE LOGITS
    ственных
    0.07
    _fonts
    0.06
     occupy
    0.06
    hic
    0.06
    _rows
    0.06
     %,
    0.06
    lack
    0.06
    .Local
    0.06
     firing
    0.06
     Stateless
    0.06
    Activations Density 0.011%

    No Known Activations