© 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
    2. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    3. Llama3.1-8B-IT
    4. Resid Post - 131k
    5. 11-RESID-POST-AA
    6. 41253
    Prev
    Next
    INDEX
    Explanations

    ,

    np_max-act · gemini-2.0-flash

    The neuron fires on the short all-caps letter sequences that appear in the header lines of a court opinion (e.g. two-letter uppercase abbreviations and acronyms like “AP,” “AM,” “F,” the commas after dates, etc.).

    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
     面
    -0.07
     дод
    -0.07
     limb
    -0.06
    شاه
    -0.06
    육
    -0.06
     Punch
    -0.06
     Jump
    -0.06
    üyle
    -0.06
    Look
    -0.06
    فهوم
    -0.06
    POSITIVE LOGITS
    يلي
    0.06
     designers
    0.06
     crowds
    0.06
    เพลง
    0.06
     Legisl
    0.06
    кадем
    0.06
    rok
    0.06
    _repeat
    0.06
     political
    0.06
    اجات
    0.05
    Activations Density 0.010%

    No Known Activations