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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & 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. 84962
    Prev
    Next
    INDEX
    Explanations

    should

    np_max-act · gemini-2.0-flash

    references to personal experiences and reflections on creativity and entrepreneurship.

    oai_token-act-pair · gpt-4o-miniTriggered by @xinyanhu8

    This neuron detects the modal “should” when expressing obligation or regret.

    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
    /;↵
    -0.08
    信
    -0.07
    (visitor
    -0.06
    _PENDING
    -0.06
    만남
    -0.06
     Angry
    -0.06
    قية
    -0.06
    ющими
    -0.06
    	assertNotNull
    -0.06
    naments
    -0.06
    POSITIVE LOGITS
    especially
    0.07
    ateurs
    0.06
     should
    0.06
     collective
    0.06
     dois
    0.06
    turn
    0.06
    PL
    0.06
     boto
    0.06
    QT
    0.06
     sequel
    0.06
    Activations Density 0.016%

    No Known Activations