© 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. 757
    Prev
    Next
    INDEX
    Explanations

    names and places

    np_max-act · gemini-2.0-flash

    The neuron is primarily triggered by personal names, especially those of political figures.

    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
     leggings
    -0.07
    -disc
    -0.07
     перед
    -0.06
     sometimes
    -0.06
     अच
    -0.06
    esteem
    -0.06
    348
    -0.06
     شي
    -0.06
     Beirut
    -0.06
     Toilet
    -0.06
    POSITIVE LOGITS
    atitude
    0.07
     Sk
    0.07
    \",↵
    0.07
     sk
    0.06
     Stability
    0.06
    hape
    0.06
     ive
    0.06
    edList
    0.06
     Mel
    0.06
    είου
    0.06
    Activations Density 0.023%

    No Known Activations