Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    3. Llama3.1-8B-IT (Instruct)
    4. Resid Post - 131k
    5. 11-RESID-POST-AA
    6. 87027
    Prev
    Next
    INDEX
    Explanations

    don

    np_max-act · gemini-2.0-flash

    words and phrases related to violent, destructive, or unethical actions and intentions.

    oai_token-act-pair · claude-3-5-sonnet-20240620Triggered by @andyrdt
    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
    No Configuration Found
    Embeds
    IFrame
    Link
    •Misaligned Persona
    Evil Persona

    No Comments

    Negative Logits
    tool
    -0.07
     кле
    -0.06
    .angle
    -0.06
     bladder
    -0.06
     jeep
    -0.06
     чет
    -0.06
    τισ
    -0.06
    .look
    -0.06
    _GRID
    -0.06
     polish
    -0.06
    POSITIVE LOGITS
    _updates
    0.07
    �
    0.07
    	uint
    0.07
     likes
    0.07
    ;&#
    0.07
     Libraries
    0.07
    	conn
    0.06
    (Initialized
    0.06
     Zub
    0.06
    jure
    0.06
    Activations Density 0.048%

    No Known Activations