© 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. 89915
    Prev
    Next
    INDEX
    Explanations

    first-person pronouns

    np_max-act · gemini-2.0-flash

    The neuron fires on the assistant’s self‐referential “I” (first-person singular pronoun) in its responses.

    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
    marshal
    -0.07
    essay
    -0.06
     cabin
    -0.06
    илось
    -0.06
    athers
    -0.06
     Luz
    -0.06
    Dick
    -0.06
    hel
    -0.06
     cows
    -0.06
    beginTransaction
    -0.05
    POSITIVE LOGITS
    /ext
    0.07
    Shock
    0.07
     가진
    0.07
    .Color
    0.07
    bff
    0.07
    ulture
    0.06
    .vert
    0.06
     Advertisement
    0.06
     �
    0.06
     आध
    0.06
    Activations Density 0.008%

    No Known Activations