© 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
    7. Resid Post - 131k
    9. 11-RESID-POST-AA
    11. 56010
    Prev
    Next
    INDEX
    Explanations

    Hello

    np_max-act · gemini-2.0-flash

    This neuron fires on greeting tokens—especially “Hello” (and its accompanying punctuation)—at the start of a text.

    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
    bed
    -0.07
    -def
    -0.07
    .has
    -0.07
    -store
    -0.07
     sce
    -0.07
    实施
    -0.06
     Betting
    -0.06
     Span
    -0.06
    Bed
    -0.06
     Бер
    -0.06
    POSITIVE LOGITS
    Proveedor
    0.06
     debug
    0.06
     huyện
    0.06
     "["
    0.06
     Bella
    0.06
    asmus
    0.06
     Doyle
    0.06
     Downs
    0.06
    Dear
    0.06
     greeted
    0.06
    Activations Density 0.003%

    No Known Activations