© 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
    2. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    3. Llama3.1-8B-IT
    4. Resid Post - 131k
    5. 11-RESID-POST-AA
    6. 9012
    Prev
    Next
    INDEX
    Explanations

    ,

    np_max-act · gemini-2.0-flash

    The neuron picks out standalone heading or label words (e.g. “Location,” “Attention,” etc.) that mark section titles in listing-style 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
     kc
    -0.07
    ари
    -0.07
     اسر
    -0.06
    締
    -0.06
    /user
    -0.06
    /colors
    -0.06
    .True
    -0.06
    され
    -0.06
    тар
    -0.06
    üc
    -0.06
    POSITIVE LOGITS
    (JNIEnv
    0.06
    Sp
    0.06
     injust
    0.06
    Skip
    0.06
     trest
    0.06
    =cut
    0.06
    Filed
    0.06
     нос
    0.06
     Phát
    0.06
    ;}↵↵
    0.06
    Activations Density 0.170%

    No Known Activations