© 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. 7-RESID-POST-AA
    11. 29109
    Prev
    Next
    INDEX
    Explanations

    Code and bots

    np_max-act · gemini-2.0-flash

    the assistant’s turn marker, especially when responding in roleplay or sexually explicit contexts.

    oai_token-act-pair · gpt-5Triggered by @vetterc0
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    andyrdt/saes-llama-3.1-8b-instruct/resid_post_layer_7/trainer_1
    Dataset (Dashboard)
    Various
    Features
    131,072
    Data Type
    float32
    Hook Name
    blocks.7.hook_resid_post
    Architecture
    standard
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     malaria
    -0.07
    одерж
    -0.06
    ยนแปลง
    -0.06
    =context
    -0.06
     kod
    -0.06
     konum
    -0.06
    Ơ
    -0.06
     bonded
    -0.06
     uyg
    -0.06
    illin
    -0.06
    POSITIVE LOGITS
    <c
    0.07
     작
    0.06
    (dto
    0.06
    Application
    0.06
     viele
    0.06
     выс
    0.06
     Devil
    0.06
     produkt
    0.06
    SECOND
    0.06
    -ST
    0.06
    Activations Density 0.042%

    No Known Activations