© 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
    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. 6121
    Prev
    Next
    INDEX
    Explanations

    coding

    np_max-act · gemini-2.0-flash

    This neuron fires on first‐person “I” plus adjacent words expressing desire or necessity (e.g. “I want,” “I need,” “where can I put…”), i.e. user requests or instructions.

    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
     gian
    -0.07
    Queue
    -0.07
    .template
    -0.07
    AEA
    -0.06
     Vale
    -0.06
     NAV
    -0.06
    /domain
    -0.06
    	Page
    -0.06
     challenges
    -0.06
     cartel
    -0.06
    POSITIVE LOGITS
     jackets
    0.07
    自分の
    0.06
     ASAP
    0.06
     mới
    0.06
     <!--<
    0.06
    .RED
    0.06
     *(*
    0.06
    ulance
    0.06
    .fullName
    0.06
    reich
    0.06
    Activations Density 0.112%

    No Known Activations