© 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
    2. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    3. Llama3.1-8B-IT
    4. Resid Post - 131k
    5. 11-RESID-POST-AA
    6. 42853
    Prev
    Next
    INDEX
    Explanations

    structures

    np_max-act · gemini-2.0-flash

    This neuron activates on uppercase initialisms or acronyms (sequences of capital letters).

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

    finds tokens that appear in message metadata or structural/header text (e.g., system/user/assistant markers, titles, and section-heading/metadata words).

    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
     cry
    -0.06
     Marg
    -0.06
     लग
    -0.06
    aginator
    -0.06
    hor
    -0.06
    名称
    -0.06
    ancing
    -0.06
    .onView
    -0.06
     Cookie
    -0.06
    (Float
    -0.06
    POSITIVE LOGITS
    طب
    0.07
    имв
    0.06
    Converted
    0.06
     считается
    0.06
    void
    0.06
    subj
    0.06
    ดาว
    0.06
    .getBody
    0.06
     helf
    0.06
     HVAC
    0.06
    Activations Density 0.663%

    No Known Activations