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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    5. Llama3.1-8B-IT (Instruct)
    7. Resid Post - 131k
    9. 11-RESID-POST-AA
    11. 129643
    Prev
    Next
    INDEX
    Explanations

    .

    np_max-act · gemini-2.0-flash

    This neuron activates on abbreviated legal titles and professional-designation tokens (e.g. “U.S.”, “Atty.”) in the attorney‐listing lines.

    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
    まず
    -0.07
    structors
    -0.06
    _TRANSFORM
    -0.06
    -messages
    -0.06
    appeared
    -0.06
     сек
    -0.06
    >"
↵
    -0.06
    pecially
    -0.06
    そんな
    -0.06
    ,k
    -0.06
    POSITIVE LOGITS
     xls
    0.06
     Lips
    0.06
    约
    0.06
    านคร
    0.06
    ;d
    0.06
     Customs
    0.06
     Marks
    0.06
     Country
    0.06
     Roth
    0.06
    реш
    0.06
    Activations Density 0.003%

    No Known Activations