Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    3. Llama3.1-8B-IT (Instruct)
    4. Resid Post - 131k
    5. 11-RESID-POST-AA
    6. 40445
    Prev
    Next
    INDEX
    Explanations

    code

    np_max-act · gemini-2.0-flash

    This neuron detects the “Cause:” and “Effect:” labels (and related instruction tokens) used when extracting cause-and-effect pairs.

    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
    _FRAMEBUFFER
    -0.07
    aviest
    -0.07
    _INITIALIZ
    -0.07
    urrency
    -0.07
    499
    -0.06
    闻
    -0.06
    kd
    -0.06
    _backend
    -0.06
    005
    -0.06
    وق
    -0.06
    POSITIVE LOGITS
     sei
    0.07
     sorumlu
    0.07
     guarda
    0.07
     alors
    0.06
     accr
    0.06
     RE
    0.06
    [S
    0.06
    bot
    0.06
     Holds
    0.06
    ,status
    0.06
    Activations Density 0.052%

    No Known Activations