© 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
    7. Resid Post - 131k
    9. 11-RESID-POST-AA
    11. 30638
    Prev
    Next
    INDEX
    Explanations

    was

    np_max-act · gemini-2.0-flash

    The neuron detects words that establish historical or contextual framing—terms specifying time periods, events, and cultural settings.

    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
    íny
    -0.07
     dipped
    -0.06
    Wa
    -0.06
    elidir
    -0.06
    ум
    -0.06
    يع
    -0.06
    _NT
    -0.06
     Ikea
    -0.06
    tors
    -0.06
     letzten
    -0.06
    POSITIVE LOGITS
    	gl
    0.07
     krb
    0.07
     TForm
    0.07
     жовт
    0.07
     hairst
    0.06
     Lifestyle
    0.06
    -east
    0.06
     �
    0.06
    _DATA
    0.06
    shopping
    0.06
    Activations Density 0.034%

    No Known Activations