© 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. 9172
    Prev
    Next
    INDEX
    Explanations

    id

    np_max-act · gemini-2.0-flash

    The neuron activates on tokens referring to personal identification documents (e.g., “ID,” “Photo,” “Driver’s license,” “identification”).

    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
     point
    -0.07
     XIII
    -0.06
    Ve
    -0.06
    .threshold
    -0.06
     Wage
    -0.06
     scratching
    -0.06
     anniversary
    -0.06
    memo
    -0.06
    вет
    -0.06
    det
    -0.06
    POSITIVE LOGITS
     huz
    0.07
    .payment
    0.07
     PROVID
    0.07
    _ENCOD
    0.07
    _define
    0.06
     محدود
    0.06
    луж
    0.06
     squeez
    0.06
     Detailed
    0.06
     repro
    0.06
    Activations Density 0.006%

    No Known Activations