© 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. 126666
    Prev
    Next
    INDEX
    Explanations

    equal sign

    np_max-act · gemini-2.0-flash

    This neuron fires on code identifier tokens—especially Java type and variable names (e.g. class names, object names, and variable identifiers).

    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
    のみ
    -0.07
     Craw
    -0.07
    Daily
    -0.06
    кін
    -0.06
    ерина
    -0.06
     اسر
    -0.06
    Organization
    -0.06
    gui
    -0.06
    函
    -0.06
     abc
    -0.06
    POSITIVE LOGITS
    ’est
    0.07
    .setX
    0.07
     robbed
    0.07
     subtree
    0.07
     barcode
    0.07
     uống
    0.06
     deficiencies
    0.06
    istory
    0.06
    'est
    0.06
     nominees
    0.06
    Activations Density 0.030%

    No Known Activations