© 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. 85827
    Prev
    Next
    INDEX
    Explanations

    punctuation

    np_max-act · gemini-2.0-flash

    The neuron primarily activates on numeric tokens (sequences of digits) in the 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
     Preston
    -0.06
    .swt
    -0.06
    moire
    -0.06
    lington
    -0.06
    poser
    -0.06
    laps
    -0.06
     Walker
    -0.06
    Ber
    -0.06
    uning
    -0.06
     _(
    -0.06
    POSITIVE LOGITS
    .ImageIcon
    0.07
    .unlock
    0.07
    Credits
    0.07
     instantiated
    0.07
    ования
    0.07
     боя
    0.07
    -photo
    0.07
    زان
    0.06
    活
    0.06
    AD
    0.06
    Activations Density 0.043%

    No Known Activations