© 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. 89115
    Prev
    Next
    INDEX
    Explanations

    Text

    np_max-act · gemini-2.0-flash

    This neuron fires on Kivy interface definitions—i.e. lines declaring Kivy layout widgets or their properties (like BoxLayout, GridLayout, size_hint, spacing, id, etc.).

    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
     climbs
    -0.06
    yes
    -0.06
    von
    -0.06
     von
    -0.06
    middleware
    -0.06
     purely
    -0.06
     marginBottom
    -0.06
    ysa
    -0.06
    ssue
    -0.06
     bows
    -0.06
    POSITIVE LOGITS
     firearms
    0.07
     Blockly
    0.07
     získ
    0.07
     çocu
    0.07
     getattr
    0.06
    .text
    0.06
    /dir
    0.06
    /Open
    0.06
     harmon
    0.06
     dokun
    0.06
    Activations Density 0.014%

    No Known Activations