© 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. 30921
    Prev
    Next
    INDEX
    Explanations

    Taking notes/remembering

    np_max-act · gemini-2.0-flash

    The neuron fires on phrases used to signal cautions or reminders (e.g. “keep in mind,” “note that,” “burn it into your memories”).

    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
     hike
    -0.07
     Bend
    -0.07
     affirm
    -0.07
     talk
    -0.07
     ==
    -0.07
    Material
    -0.06
    hands
    -0.06
     opportunity
    -0.06
     files
    -0.06
     Judges
    -0.06
    POSITIVE LOGITS
     pulver
    0.06
    etable
    0.06
    349
    0.06
    ARATION
    0.06
     Auschwitz
    0.06
    owania
    0.06
     registrazione
    0.06
    forgettable
    0.06
    PRESSION
    0.06
    +b
    0.05
    Activations Density 0.020%

    No Known Activations