© 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 EvalsExportsinterp-engineNEWAPI 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. 27-RESID-POST-AA
    11. 12445
    Prev
    Next
    INDEX
    Explanations

    2

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    andyrdt/saes-llama-3.1-8b-instruct/resid_post_layer_27/trainer_1
    Dataset (Dashboard)
    Various
    Features
    131,072
    Data Type
    float32
    Hook Name
    blocks.27.hook_resid_post
    Architecture
    standard
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    .AreEqual
    -0.06
    =settings
    -0.06
    叫
    -0.06
     ”↵
    -0.06
    Comput
    -0.06
     Tonight
    -0.06
     ủy
    -0.06
    	NSString
    -0.06
     lg
    -0.06
    舍
    -0.06
    POSITIVE LOGITS
     、
    0.07
     contradiction
    0.07
    uteur
    0.07
    Order
    0.07
     aks
    0.07
     wages
    0.07
     Heal
    0.06
    렸다
    0.06
    label
    0.06
     carte
    0.06
    Activations Density 0.004%

    No Known Activations