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. 130196
    Prev
    Next
    INDEX
    Explanations

    graphics

    np_max-act · gemini-2.0-flash

    This neuron activates on mentions of graphics cards (especially Spanish terms like “tarjeta gráfica,” “gráfica,” and specific GPU model names).

    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
     window
    -0.08
     Klan
    -0.06
    LE
    -0.06
     жал
    -0.06
    xia
    -0.06
     Hal
    -0.06
    .Intent
    -0.06
    ofi
    -0.06
     Eduardo
    -0.06
     Zhang
    -0.06
    POSITIVE LOGITS
    _RESET
    0.06
     ;
    ↵
    ↵
    0.06
    check
    0.06
    USER
    0.06
     Fixture
    0.06
    '"
    0.06
     ;-
    0.06
     respiratory
    0.06
     reverted
    0.06
    .'),↵
    0.06
    Activations Density 0.017%

    No Known Activations