© 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. 68647
    Prev
    Next
    INDEX
    Explanations

    Location and Halting

    np_max-act · gemini-2.0-flash

    This neuron activates on verbs and phrases that signal ending or termination—e.g. “stop,” “end,” “ending,” or “blur just short of the edge.”

    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
    ([]
    -0.07
     covert
    -0.07
    کات
    -0.06
    (empty
    -0.06
     doctrine
    -0.06
    .Entity
    -0.06
    页面
    -0.06
    sg
    -0.06
    .mkdirs
    -0.06
    	swap
    -0.06
    POSITIVE LOGITS
     Xu
    0.07
    abay
    0.07
    avatar
    0.06
     تحص
    0.06
    SECTION
    0.06
     �
    0.06
     accol
    0.06
     joe
    0.06
     cameras
    0.06
     endDate
    0.06
    Activations Density 0.013%

    No Known Activations