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

    Neuronpedia

    Jacobian LensNEW
    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. 19-RESID-POST-AA
    11. 110626
    Prev
    Next
    INDEX
    Explanations

    talking

    np_max-act · gemini-2.0-flash

    The neuron detects self-referential/reflexive language—tokens and phrases that refer back to the subject itself (e.g., "self", "itself", "talking to itself", "self-...").

    oai_token-act-pair · gpt-5-miniTriggered by @yooniel31

    self-reference

    np_max-act · gpt-5-miniTriggered by @yooniel31
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    andyrdt/saes-llama-3.1-8b-instruct/resid_post_layer_19/trainer_1
    Dataset (Dashboard)
    Various
    Features
    131,072
    Data Type
    float32
    Hook Name
    blocks.19.hook_resid_post
    Architecture
    standard
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     mirrors
    -0.07
     Since
    -0.07
    コ
    -0.06
    Since
    -0.06
     variants
    -0.06
     customers
    -0.06
    itation
    -0.06
    "There
    -0.06
     do
    -0.06
    ۱۴
    -0.06
    POSITIVE LOGITS
    ungkin
    0.07
    َح
    0.07
     �
    0.06
    #error
    0.06
    bitmap
    0.06
     unsett
    0.06
     Pathfinder
    0.06
    Tot
    0.06
    Jimmy
    0.06
     olduğ
    0.06
    Activations Density 0.234%

    No Known Activations