© 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 (Instruct)
    7. Resid Post - 131k
    9. 11-RESID-POST-AA
    11. 44303
    Prev
    Next
    INDEX
    Explanations

    :

    np_max-act · gemini-2.0-flash

    The neuron fires primarily on the imperative cue “say,” i.e. detecting when the user is being instructed to say something.

    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
     eve
    -0.07
     Ger
    -0.06
    /index
    -0.06
    (surface
    -0.06
    /session
    -0.06
    肩
    -0.06
    itelist
    -0.06
    (Client
    -0.06
     Surely
    -0.06
     sewer
    -0.06
    POSITIVE LOGITS
     предназнач
    0.07
     titten
    0.07
     slump
    0.06
     đàn
    0.06
     فضای
    0.06
     bbox
    0.06
    (inode
    0.06
    쳐
    0.06
    _qos
    0.06
     своїм
    0.06
    Activations Density 0.002%

    No Known Activations