© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 6193
    Prev
    Next
    INDEX
    Explanations

    do

    np_max-act · gemini-2.0-flash

    The neuron activates on the auxiliary verb “do” when it appears in first-person assertions (e.g. the “do” in “I do not…”).

    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
     alternatives
    -0.07
    Verts
    -0.07
    contrast
    -0.07
    -0.06
    벌
    -0.06
    ements
    -0.06
     unsus
    -0.06
     avoir
    -0.06
    brands
    -0.06
     Canucks
    -0.06
    POSITIVE LOGITS
     did
    0.13
     do
    0.11
     does
    0.10
     DO
    0.09
     DID
    0.09
     Do
    0.09
    Do
    0.07
     billing
    0.07
     Did
    0.07
    Did
    0.07
    Activations Density 0.032%

    No Known Activations