© 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. 89328
    Prev
    Next
    INDEX
    Explanations

    a

    np_max-act · gemini-2.0-flash

    This neuron detects the indefinite article “a” when it introduces a technical component or unit (i.e. the start of a patent‐style component description).

    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
     Род
    -0.07
     Levine
    -0.06
    (edges
    -0.06
     Stard
    -0.06
     کود
    -0.06
    Were
    -0.06
    .ast
    -0.06
     Εν
    -0.06
     बन
    -0.06
     downloads
    -0.06
    POSITIVE LOGITS
    iona
    0.07
     kurtar
    0.07
    ’na
    0.07
     meisten
    0.06
     McA
    0.06
     objet
    0.06
    ема
    0.06
    、ア
    0.06
    ueva
    0.06
     sap
    0.06
    Activations Density 0.016%

    No Known Activations