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

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI 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. 58619
    Prev
    Next
    INDEX
    Explanations

    when

    np_max-act · gemini-2.0-flash

    The neuron responds to temporal connectors and sequencing words (e.g. “when,” “had,” “report,” “submit”) that mark clause boundaries and time relationships.

    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
    .fxml
    -0.08
     songs
    -0.07
     cush
    -0.07
     stares
    -0.07
    (Util
    -0.06
    ्रथ
    -0.06
    LAR
    -0.06
    am
    -0.06
    Authorities
    -0.06
    Range
    -0.06
    POSITIVE LOGITS
     зал
    0.07
     지금
    0.06
     قسم
    0.06
    /Public
    0.06
     Margin
    0.06
    rud
    0.06
    观看
    0.06
     Č
    0.06
     sensed
    0.06
    /css
    0.06
    Activations Density 0.006%

    No Known Activations