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

    Neuronpedia

    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. 11-RESID-POST-AA
    11. 128360
    Prev
    Next
    INDEX
    Explanations

    comma

    np_max-act · gemini-2.0-flash

    The neuron activates on individual food‐related ingredients (e.g., flour, tomatoes, chicken, lentils), indicating it’s detecting ingredient names in a recipe or pantry list.

    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
     تا
    -0.06
    jím
    -0.06
    ๋
    -0.06
    olkata
    -0.06
    Ymd
    -0.06
     pizza
    -0.06
    Sugar
    -0.06
     burgers
    -0.06
    ORIES
    -0.06
     spanking
    -0.06
    POSITIVE LOGITS
     generics
    0.07
     출시
    0.06
     створ
    0.06
    _maps
    0.06
    .name
    0.06
    _group
    0.06
    /ng
    0.06
     mixins
    0.06
     získ
    0.06
    	Debug
    0.06
    Activations Density 0.018%

    No Known Activations