© 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. 124904
    Prev
    Next
    INDEX
    Explanations

    vulnerable

    np_max-act · gemini-2.0-flash

    The neuron fires on mentions of vulnerable or at‐risk demographic groups (e.g. “vulnerable populations,” “children,” “elderly”) when discussing hazards or protections.

    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
     noticeable
    -0.07
     좋은
    -0.07
    スカ
    -0.06
     Mavericks
    -0.06
    説
    -0.06
     matchup
    -0.06
     GOODMAN
    -0.06
     उम
    -0.06
    693
    -0.06
    ього
    -0.06
    POSITIVE LOGITS
     Carn
    0.07
    _hal
    0.07
    packageName
    0.06
     Bas
    0.06
    بع
    0.06
     -(
    0.06
    (category
    0.06
     imread
    0.06
    (transform
    0.06
     Quint
    0.06
    Activations Density 0.011%

    No Known Activations