© 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. 109406
    Prev
    Next
    INDEX
    Explanations

    who, what, where, when, why, how

    np_max-act · gemini-2.0-flash

    The neuron is detecting the “question‐word” tokens (who, what, when, where, why, how) in the 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
    _INSTALL
    -0.07
     LEVEL
    -0.06
    Вт
    -0.06
    Tro
    -0.06
    _INV
    -0.06
    块
    -0.06
     timestep
    -0.06
    .Pop
    -0.06
     bốn
    -0.06
     kr
    -0.06
    POSITIVE LOGITS
     اك
    0.07
     TableView
    0.06
    設定
    0.06
     gentleman
    0.06
     Scala
    0.06
     dạng
    0.06
     сом
    0.06
     색
    0.06
    (attrs
    0.06
     mana
    0.06
    Activations Density 0.007%

    No Known Activations